Jasa Konsultasi Disertasi Keuangan (1) dan Perbankan S3, hubungi WA 0821 2230 7021: Contoh Tesis mengenai OFDM modulation

Studyof OFDM modulation

Eldo Mabiala, Mathias Coinchon

Eurecom institute

Supervisor: Karim Maouche

Date: december 1999

Semester ro ect

Table des matières

1 Introduction 1

1.1 Aim of project 1

1.2 Abstract 1

1.3 Plan of report 2

2 OFDM 3

2.1 Channel 3

2.1.1 Frequency selectivity 3

2.1.2 Delay spread 3

2.2 Qualitative description 4

2.2.1 The problem of wideband transmission on a single carrier 4

2.2.2 Multicarrier transmission 4

2.2.3 Orthogonality 5

2.2.4 The guard interval 6

2.2.5 Comparison with other schemes 6

2.3 Mathematical description 6

2.3.1 Multicarrier approach: 6

2.3.2 Inverse Fourier Transform: 7

2.3.3 Adding a guard interval: 7

2.3.4 Limitation of bandwidth: 8

2.3.5 Channel and receiver parts: 9

2.4 Synoptic of an OFDM transmission system 10

2.5 Problems in OFDM 11

2.5.1 Orthogonality: 11

2.5.2 Synchronization: 12

2.5.3 Peak Average Power Ratio (PAPR): 12

2.5.4 Effect of clipping in OFDM signals: 13

2.5.5 Phase noise: 13

2.5.6 Frequency error: 13

2.5.7 Conclusion 14

2.6 Design of an OFDM system 14

2.6.1 Bandwidth 14

2.6.2 Number of carriers 14

2.6.3 Guard interval 15

2.6.4 Modulation 15

2.6.5 Coding 15

3 Applications 16

3.1 General 16

3.2 DAB 16

3.2.1 What bandwidth to use for OFDM? 16

3.2.2 How many carriers could we use? 17

3.3 ADSL 18

3.3.1 HDSL 18

3.4 Hiperlan/2 19

3.4.1 Definition 19

3.4.2 What is the mission statement of H2GF? 19

3.4.3 Who are the founding members? 19

3.4.4 When will the H2GF be launched? 19

3.4.5 How does HiperLAN2 work? 19

3.4.6 Examples of use and scenarios: 20

3.5 Other applications 20

4 Simulation 21

4.1 Results 21

4.1.1 OFDM QAM transmission with only additive noise 21

4.1.2 OFDM QAM transmission with Raleigh channel model 22

4.1.3 OFDM BPSK transmission 22

4.1.4 BPSK transmission on one carrier 23

4.1.5 OFDM 4-PAM transmission 23

4.1.6 4-PAM Modulation one carrier 24

5 Conclusion 25

5.1 Future 25

5.1.1 Multi Carrier CDMA (MC-CDMA): 25

5.2 Conclusion 25

6 Annexes 27

6.1 Bibliography : 27

6.2 Glossary 28

6.3 Matlab code 28

6.3.1 OFDM simulator Error! Bookmark not defined.

6.3.2 Function QAM.m Error! Bookmark not defined.

6.3.3 Function gene_fad Error! Bookmark not defined.

6.3.4 Function mod_fad Error! Bookmark not defined.

6.3.5 Function propag.c Error! Bookmark not defined.

1 Introduction

General presentation andplan ofreport.

1.1 Aim of project

The aim of the project is the study of the Orthogonal Frequency Division Multiplexing modulation. The questions are what are the main principles that technology use ? What are the advantages and disadvantages in regard with the other schemes ? In which case, environment is it used ? what is the actual state-of-the art of it ? Where do we meet OFDM technology nowadays, and in the future ? What are today’s standards ? Is it a mature technology

1.2 Abstract

Orthogonal Frequency Division Multiplexing (OFDM) is a new digital modulation technique who consists of transmitting a data stream on several carriers instead of using only one carrier. The general concept has been first introduced in 1971, but it’s only in the last decade, with the development of Digital Signal Processors (DSP) that applications become visible.

OFDM is mainly used on wideband transmissions. We’ll see that OFDM is well suited for transmissions in frequency selective channels. Such a situation is met for example in multipath environments.

Applications nowadays find place in most of high data rate, wideband transmissions.

• In audio and television broadcasting (DAB and DVB) where usually high data rate transmissions are required in multipath environment. We’ll see too that OFDM permits the use of a single frequency network (SFN) where one broadcast multiplex can be sent at the same frequency with several transmitters.

• In wireline transmission: ADSL to transmit data at a very high rate on copper wire lines. The problem here is that characteristics of the line are not known and may change among users.

• In wireless LAN networks with HiperLAN/2 standard where transmissions occurs at very high data rate (~10 Mb/s) in indoors environment (strong multipath environment).

1.3 Plan of report

In the next chapter, we’ll explain the principle of OFDM modulation by introducing things intuitively. We’ll then see some points more in details by doing a mathematical description. Then, we’ll focus more on the practical aspects of OFDM : the implementation and the problems that modulation suffer from.

The third chapter is dedicated to the main applications of OFDM nowadays. We’ll focus particulary on Digital Audio Broadcasting (DAB) application since it covers most aspects of an OFDM transmission.

Also, in order to touch the problem of implementation, we’ll make an OFDM transmission simulator on MATLAB.

2 OFDM

Description of Orthogonal Frequency Division

Multiplexing modulation.

2.1 Channel

When doing radio transmission on high frequencies (VHF and higher) we are often confronted to a multipath environment. Such environment are found mostly in urban areas where buildings reflects waves.

2.1.1 Frequency selectivity

One problem of multipath is that the resultant of waves from different paths can be constructive or destructive depending on the position, so signal change over time when moving. You can experiment that when listening on FM radios while driving in a city: signal is cut by noise when moving.

Generally we speak of a frequency selective channel. Characteristics may change fast when moving. That’s why the channel is also time varying.

Typicalfrequency response of a channel sufferingfrom mutlipath propagation

The aim of equalization is to compensate the problems introduced by frequency selectivity.

2.1.2 Delay spread

Another effect that affect digital transmission, is that the signal coming from

different paths has different time delays depending on the length of path. A

consequence of that is memory of channel which cause interference between symbol received (ISI).

2.2 Qualitative description

2.2.1 The problem of wideband transmission on a single carrier

When transmitting wideband on frequency selective channels, equalisation must be performed in order to avoid intersymbol interference. Equalisation try to make the channel flat. In order to do that channel state information is needed. Training sequences have then to be transmitted periodically to estimate channel. Channel estimation is performed by several calculations and is then CPU time consuming. So when data rate is high and when characteristics of channel change rapidly CPU power needed is high and system become expensive.

Also is the symbol period very small compared to the channel memory. That cause strong Intersymbol Interference (ISI) and equalization is needed to correct that problem.

0)o f

2.2.2 Multicarrier transmission

The idea of multicarrier transmission is to divide bandwidth in several narrow band transmissions so that the channel looks flat on each carrier. The data stream to transmit is then split among the carriers instead of being tranmsitted on one carrier with large signal bandwidth. That’s what is meant by Frequency Division Multiplex (FDM). The advantage is that no or less complex equalization is needed.

The symbol period on each carrier become large and the effect of channel memory (length of channel impulse response) become less destructive on the symbols, so intersymbol interference (ISI) is reduced and less equalisation is needed.

Each carrier is modulated using any known modulation scheme like QAM and PSK schemes.

Reexp(0)0t+ gyp)}

2.2.3 Orthogonality

In a transmission system, we want the occupied bandwidth on the channel to be as small as possible. For that, in a multicarrier system, we try to set a minimum frequency space between carriers without having intercarrier interference (ICI). The minimum space is reached when carriers are orthogonal to each other, signal from each can have a small overlap on the other without causing interference. That’s what is meant by the « O » (Orthogonal) of OFDM. We’ll see further that the inverse fourier transform has that orthogonality property.

1/ T

2.2.4 The guard interval

One way to avoid Inter-symbol interference is to set a small gap equal to the duration of delay spread between the symbols. So, each symbol does not affect the next one. We’ll also see later that this interval plays an important role in the implementation.

2.2.5 Comparison with other schemes

Simple schemes (QAM, ... ) as we have just seen, suffer from strongISI when channel memory increase and then strong equalisation is needed. Also the risk of signal loss is high in frequency selective fading channels

CDMA (or spread spectrum) schemes. In order to have high data rate transmissions, receiver has to compute correlations at a high rate and it is CPU time consuming.

2.3 Mathematical description

Let us consider the general problem of transmitting a signal S(t) over the time-varying channel c(t,T). The sampling version of S(t) is a bit stream S[n] that is demultiplexed onto N carriers.

2.3.1 Multicarrier approach:

The bit stream S[n] is shared onto all carriers, each one to be produced some data to transmit. Then the bit stream is divided in sub-stream (Xn [k0<_k<_N_1 called OFDM symbol, we use two indexes n being time index and k the sub-stream index (index used for carrier recognition). The carriers do not receive an amount of bit since the bit stream is pre modulated, the required amplitude and phase of the carrier is then calculated using BPSK, QPSK or QAM before demultiplexing. A variant of OFDM is COFDM (coded orthogonal frequency division multiplexing), where forward error coding is applied to the signal before transmission, to overcome errors due to lost carriers from frequency selective fading, channel noise and other effects of propagation. But the main focus of this project is OFDM.

Xn [0]

Xn [1]

Xn [ N-1]

2.3.2 Inverse Fourier Transform:

OFDM uses the available spectrum efficiently by spacing the channels much closer together. This is achieved by making all the carriers orthogonal to one another, preventing interference between the closely spaced carriers. To generate OFDM successfully the relationship between all carriers must be carefully to maintain the orthogonality of the carriers. For that, after choosing the spectrum required, we have to convert it back to its time domain signal using an Inverse Fourier Transform. In most applications, an Inverse Fast Fourier Transform is used, it performs the transformation very efficiently, and provides a simple way of ensuring the carrier signals produced are orthogonal.

Xn[0] Xn[1]

Xn[ N-1] IFFT xn[0]

xn [1]

xn [N-1]

1 N k

Xkexp j 2 l

N

k_0 N

-1

n _ n

For l = 0, 1, ..., N-1

2.3.3 Adding a guard interval:

One of the most important properties of OFDM transmissions is the robustness against multipath delay spread. This is achieved by having a long symbol period, which minimizes the inter-symbol interference. The level of robustness, can in fact be increased even more by addition of a guard interval between transmitted symbols. The guard period allows time for multipath signals from the previous

symbol to die away before the information from the current symbol gathered. The most effective guard interval to use is a cyclic extension of the symbol. Why?

The Fast Fourier Transform that we will use at the receiver, transforms a cyclic time domain signal into its equivalent frequency spectrum. The signal xn k is not

necessary cyclic... Let form a cyclic signal of N+L-1 samples by repeating the last L-1 samples xnk at the beginning of the signal. This technique is called guard

interval by cyclic prefix. The number L of samples to repeat is taken more than the memory of the channel. So we the resulting cyclic prefixed signal is:

x N  L  1 ,..., x N 1 , x 0 , x 1 ,...x N 1

n n n n n

2.3.4 Limitation of bandwidth:

The frequencies used in OFDM are regularly spaced in the spectrum required

so, the time domain signal xn k has to be limited in the frequency domain. For

this reason we will make a convolution between the signal and a window function that has a spectrum limited exactly on the required one. (All window functions have a limited spectrum). We have chosen to use the raised cosine function to limit our signal spectrum and we will note it g* t. The signal to send over the channel is:

NL 1

l 0

Example of an OFDM spectrum

2.3.5 Channel and receiver parts:

Let consider the problem of transmitting the signal xnm

~ over the time

varying linear channel ct, without additional noise. If we call cm the sampling version of the channel, then the output of obtained by the channel is:

NL 1

~ yn m xnlcm l 

~ 

l 0

m = 0,1,..., N+L-1

The receiver basically does the reverse operation to the transmitter. The signal

received is ynm

~ that has N+L-1 samples, before the demodulation, we have to drop the L last samples of the received signal, and then remove the guard period, in order to use correctly the Fourier Transform properties. Indeed the demodulation operation is a simple FFT according to the IFFT used as modulation. We need to find out N samples (one per carrier) as at the modulator input, let apply an FFT to the signal yn  m

~ .

Y~k  yl

~ exp 2jk

n n N

~    1

Y k

       

 

N 1 N L ~    

n l 

 x m c l  m exp  2  jk

n 

l  0  m  0  N

~ N  L  1 N 

       

 1 

x m c l m

~      l

Y k   exp 2

n n   jk

 m0 l0 N

~ NL1  m

Y k x m

~ jk ck

n   exp 2

 

 

n  

 m 0 N 



The first term of the multiplication above looks like a Fourier Transform expression. In fact, if we restrict the summation index from m = 0 to m = N-1 it is equivalent to drop the L last samples of the signal y~n m. So the signal produces

by the demodulator is:

~N 1

   

 m 

Y k x m

~ jk c  k

n   exp 2

  

 

n  

 m 0 N 

 

Yn k  xn kck

Yn k  Xn k Ck

A simple division by the channel frequency response gets back the transmitted

signal. This modulation does not need any equalization and the data samples are

then combined back to the same size as the original data.

2.4 Synoptic of an OFDM transmission system

Implementation of OFDM modulation nowadays is almost everytime performed digitally.

Here’s a model of an OFDM transmission system

• Data coming from the input are arranged into vectors with number ~

of components equal to the number of carriers. Each component N

is composed by a number of bits depending on the alphabet of the modulation scheme used on the next stage. For example, if we use a 1536 carriers system with BPSK, we’ll have vectors of 1536 component each one composed by 1 bit (BPSK is 2-ary).

• Each component (group of bits) is mapped into a complex symbol depending on the alphabet of the modulation scheme used. For example, with BPSK the alphabet is { -1 ; +1 }.

• In order to obtain real samples after IFFT, a 2*Number of carrier points IFFT is done with:

= X k

*,=1,...,N

k k

X = Re ( X ), X ~ = Im( X

0 0 N 0

• The Inverse Fast Fourier Transform algorithm (IFFT) is applied to the vector giving a real samples vector.

•The guard interval is added at the beginning of the vector by repeating the components of the end. Vectors are concatenated to form a time signal (parallel/serial conversion)

• Windowing the signal is necessary to limit the bandwidth. Most used window is the raised cosine.

• The signal is then passed trough the channel. Channel is modeled by a linear system with frequency response c(t) together with a source of additive Gaussian noise.

• At the reception, signal is rearranged again into vectors (serial/parallel conversion) and guard interval is dropped.

• Fast Fourier Transform (FFT) is computed in order to get back the complex vector of symbols.

2.5 Problems in OFDM

2.5.1 Orthogonality:

As seen above the fact to have several carriers is actually advantageous whenever they are mathematically orthogonal. So carriers orthogonality is a constrain that can leads to a wrong operation of OFDM systems if not respected. The orthogonality is provided by IFFT that a numerical manipulation, an error of computation could change lightly spacing between to consecutive carriers and break the orthogonality of the whole system. In this case OFDM loses all its efficiency, because the notion of orthogonality is an absolute one.

2.5.2 Synchronization:

One of the crucial problems in the receiver is to sample the incoming signal correctly. If the wrong sequence of samples is processed, the Fast Fourier Transform shall not correctly recover the received data on the carriers. The problem is more embarrassing when the receiver is switched on. There is therefore a need for acquiring timing lock. If the signal transmitted is really time domain periodic, as required for the FFT to be correctly applied, then the effect of the time displacement is to modify the phase of all carriers by a known amount. This is due to the time shift theorem in convolutional transform theory.

Symbols

Time

Sampling

However, the signal is not really repetitive, we have cheated and performed the mathematical transform as if it were repetitive, but then chosen different symbols and transmitted them one after the other. The effect of the time shift would then be not only to add the phase shift referred to above, but also to add some intersymbol interference with adjacent symbols. This interference could hardly degrade reception.

To avoid these problems, we decide to transmit more than one complete sequence of time samples in order to increase the tolerance in timing. It’s an additional data guard interval. It is built by repeating a set as long as channel memory of last samples taken in the original sequence. The longer the guard interval, the more rugged the system, but guard interval does not carry any useful information and its transmission leads to a penalty of power.

One technique used to obtain good synchronization is to add between each OFDM symbol a null (zero samples) symbol. This technique is used in DAB for time synchronization.

2.5.3 Peak Average Power Ratio (PAPR):

When the phase of different subcarriers add up to form large peaks, an important complication comes in OFDM systems. This problem is called Peak Average Power Ratio (PAPR) and it is defined for each OFDM signal on a time interval [n, n+Ts] by the following formula:

For continuous signals

For sampled signals

max

Xn = E x k

k { [ ] 1

n 2

In OFDM systems PAPR can have very high values for certain input sets of sample (Xn [k] and overload non-linear characteristics of systems, causing inter

modulations among different carriers and undesired out-of-band radiation.

Another main drawback of PAPR can be seen as quantization noise domination towards the performance of system. This domination can be excited by avoiding the clipping effect of the maximum level of the Digital to Analog Converter (DAC) that is set too high.

Various techniques are proposed to reduce PAPR in OFDM signals, but that reduction is not obvious because PAPR and SNR are closely linked. We will not expound those techniques in this paper, they can be found in reference [14].

2.5.4 Effect of clipping in OFDM signals:

When transmitted signals have high PAPRs, amplifiers may produce “clipping”. In some way, clipping can be regarded as peaks of the input signal being simply cut-off by amplifiers. Consequences of clipping are out-of-band radiation and inter symbol interference between subcarriers. In order to avoid these undesired effects that reduce OFDM performances, one has either to use amplifiers with dynamic range, or try to reduce PAPR. The first alternative is expensive, the second one is more often used.

2.5.5 Phase noise:

At the receiver, a local oscillator can add phase noise to an OFDM signal, for example. The phase noise could so have two effects those are: Common Phase Error (CPE) due to a rotation of the signal constellation and, Inter Carrier Interference (ICI), similar to additive Gaussian noise. The BBC R&D have made analysis of the effects of phase noise on an OFDM signal, this analysis shows that CPE arises simultaneously on all carriers. Indeed, the signal constellation within a given symbol is subject to the same rotation for all carriers and this effect can be corrected by using reference information within the same symbol. Unfortunately, ICI is more difficult to overcome, due to the additive noise, which is different for all carriers. This difference can be interpreted as a loss of orthogonality.

2.5.6 Frequency error:

An OFDM system can be subject to two types of frequency error. They are Frequency offset (as might be caused by the tolerance of the local oscillator frequency) and, Error in the receiver master clock frequency (which will cause the spacing of the demodulating carriers to be different from those transmitted).

Before to find solutions to those problems, the system designer needs to determine how much residual frequency error is permissible, and understand exactly how errors affect the received signal.

Both of these error situations have been analyzed so, a frequency offset affects most carriers equally, with the very edge carrier less affected. ICI resulting from a fixed absolute frequency offset increases with the number of carriers, if the system bandwidth is kept constant. About error in the receiver clock frequency, in absence of frequency offset, it affects carriers unequally (the center carrier suffers a little while the worst affected carrier lies close to, but not at, the edge).

2.5.7 Conclusion

At the reception, it is very important to distinguish the starting point of FFT to avoid wrong demodulation. And so synchronization has to be precise. It explains the use of special symbols (pilot) for synchronization in transmission.

Hardware design of transmitter and receiver is important because of high peak to average ratio which cause distorsions if dynamic range of amplifiers and converters is not high enough.

OFDM is very sensitive to carrier frequency offsets. Such offsets are mainly the cause of receiver local oscillators instability and doppler effect when mobile is moving.

2.6 Design of an OFDM system

2.6.1 Bandwidth

Occupied bandwidth is of course directly related to the data rate to transmit. However, the question is , what is the minimum bandwidth to take in order to obtain enough diversity and avoid the loss off all the signal in frequency selective fading environments. On the other hand much bandwidth means also much transmitting power. There is a tradeoff between bandwidth and transmitted power.

That optimal bandwidth is found by channel simulations and field test trials. In DAB, for example, a bandwidth of 1,5 Mhz is a good compromise for the type of propagation conditions that apply.

2.6.2 Number of carriers

We have seen that the greater the number of carriers, the greater the symbol period on each carrier and so lees equalization is needed and the greater the diversity offered by the system. However, with differential modulation, it is important that the channel not vary too much during one symbol period. This is not the case when the receiver is moving because of dopler effect and short term fading. Then a great number of carrier will limit the moving speed. This is another tradeoff of OFDM.

Another problem is the complexity in the implementation increase when carrier number increase because large FFT are needed.

To continue with the DAB example, 1536 carriers has been found to be a good compromise. That lead to a carrier spacing of 1kHz and a symbol period of 1ms. Moving speed of mobiles shouldn’t get over 160 km per hour.

2.6.3 Guard interval

The tradeoff of guard interval is to set it large enough to avoid intersymbol interference depending on the memory of channel and transmitter position spacing in a single frequency network. On the other hand, we want it to be as small as possible as it carries no information and can be seen as a spoil of bandwidth.

In wireless systems, a guard interval of 25% of symbol period is often met and seems to be a good compromise. That is the value taken for DAB, it allows a maximum distance of about 80 kilometers between transmitters

2.6.4 Modulation

The modulation scheme used on each carrier depends on the BER needs.

In DAB, QPSK is used but for higher order systems 16, 64 or 256 QAM is used.

2.6.5 Coding

Channel coding is very important in OFDM systems. When we speak of diversity,

that is possible because information is redundant among the carriers.

Coding associated with frequency (among carriers) and time interleaving make the system very robust in frequency selective fading.

In litterature, coded OFDM systems are sometime called COFDM.

3 Applications

Applications ofOFDM in digital communications

nowadays.

3.1 General

We have seen that OFDM is digital transmission technique well suited for wideband, high data rate transmissions. The main advantage is that less equalisation is necessary.

A consequence of that is that OFDM is not a very good solution for one to one communications with several users on a shared channels, because of the problem of frequency allocation. However on super high frequency bands (SHF) and Extremely high bands (EHF) where occupied bandwidth is not a great problem, OFDM may be a good solution for one to one communications. But, nowadays, OFDM is mainly used for one to many (broadcast) communications like radio or television broadcasting. That’s why we find OFDM on several new digital broadcasting systems such as DAB and DVB.

3.2 DAB

Digital Audio Broadcasting (DAB) is an international, standardized digital broadcasting system developed by the European EUREKA-147 Project. The system should completely replace, in the future, the well-known analog FM (Frequency Modulation) radio system on the 88-108 MHz frequency band.

The DAB system is digital and provides CD-like audio quality. DAB is much more robust to interferences and is well suited for mobile reception like in a car. New possibilities are available on receivers like for example multimedia features (image and texts). The transmission scheme used for DAB is OFDM modulation

DAB is being deployed around Europe and some other countries nowadays, receivers are still expensive like CD players in the beginning of eighties but we might expect to see a wide use of it in the next 5 years.

3.2.1 What bandwidth to use for OFDM?

The wider the bandwidth, the more probably that the system overcome the

correlation bandwidth of the channel.

Problem to overcome: Short delay echoes are the main problems to overcome, and as these are always present there is no hard bound. The narrower the bandwidth,

the more likely it is that the whole signal will be affected. There is a trade off between bandwidth and transmitter power.

Bandwidth: 7MHz few problems

2 MHz degradation of 1dB in performance at each point

<1.5MHz degradation starts to increase

200KHz used for FM sound, then the margin required would be an additional 6dB or so.

Trade off: 1.5MHz for the type of propagation conditions that apply to mobile and portable radio reception.

Bit-rate: On each carrier the modulation system used is QPSK, the carriers are separated by a gap of around 1/Ts, where Ts is symbol period. The maximum bit rate available is so 2bit/s/Hz of the bandwidth. This figure is reduced by the inefficiency (signal redundancy) of the guard interval, the null symbol and the error coding. For DAB, this brings the useful bit-rate down about 1 bit/s/Hz of the bandwidth.

Therefore a DAB system will provide just less than 1.5Mbit/s of useful data. This is considerably more than 256kbit/s that needed for high-quality stereophonic program, so the implication is that several broadcast programs will share the same multiplex.

3.2.2 How many carriers could we use?

About the number of carriers, the more there are, the greater is the resolution of the diversity offered by the system. For the differential demodulation to work properly, the multipath environment must change slowly from symbol by symbol. Thus, there is limit to the symbol and hence the number of carriers. For static reception, this is not a major problem. But for mobile reception, the motion of the vehicle leads to changes in the multipath environment. Over a symbol period, a vehicle moving at a velocity v m/s will travel vTs*f/c wavelengths. This is fd*Ts, where fd is the maximum Doppler shift. If this is to introduce negligible phase distortion, then the function fd must be small. A figure of fd<0.02 has been proposed as suitable for general use. To achieve high vehicle velocities (>160km/h) it has been necessary to adopt three modes of operation, each mode being suited to a difference part of VHF and UHF frequency bands. The main difference between modes is the symbol period and, as a direct consequence, the number of carriers.

If it will be only a question of single transmitter, then the significant echoes would all be relatively short. Surveys indicate that guard interval of the order of 10µs would satisfactory for the majority of location, in the UK for example. The use of several transmitters puts a limit on the minimum guard interval that should be used. The transmissions from areas that are some distance away can reach quite high levels on occasions of anomalous propagation. This has to be a compromise between a small number of high power transmitters spaced by about 50km, or a much larger number of low-power transmitters. Because the first option is likely to

be cheapest, the guard interval is to set about 250µs, equivalent to a maximum in transmission distance at the receiver.

The symbol period need not be directly related to the guard interval. It is a just question of how much of the symbol period is repeated in the guard interval. This is purely a matter of efficiency, as the power transmitted in the guard interval does not form a useful part of the data information in the receiver unless there are substantial echoes. To minimize the power loss by the system, it is desirable to keep the guard interval to as low a percentage as possible of the symbol period. In practice, a guard interval of the order of 25% of the symbol period has been found to be good compromise.

If we consider a guard interval of 250µs, this leads to a symbol period of 1ms and a carrier spacing of about 1 kHz. It means approximately 1500 carriers (W=1.5MHz) in the minimum bandwidth that would be desirable for one OFDM transmission. In practice, one uses 1536 carriers with a guard interval of 512. Then the channel receives symbols that have 2048 length so, the guard interval is the quarter of the symbol total length.

3.3 ADSL

Asymetric Digital Subscriber Line (ADSL) is a technique to transmit high data rates (up to 6 Mb/s downlink, 640kb/s uplink) on Subscriber Lines (telephone lines). Such lines consist of twisted copper wires. The idea is to use the full capacity of the line instead of using only 4 khz needed to transmit voice. Occupied bandwdth goes to 1.1 MHz.

The main problem is that the characteristics of the line change among users. They change with distance, presence of bridged taps in the line, neighbourhood of other lines, ... . The results are reflections at certain frequencies which cause attenuation, velocity dependant of the frequency which causes ISI. The situation is very similar to wireless channels.

There are 2 possible modulation schemes usable for ADSL : CAP (Carrierless amplitude phase) that is similar to QAM and Discrete Multitone (DMT) that is another appellation for OFDM.

Nowadays, it seems that DMT is the retained candidate for ADSL. The downlink consist of 222 tones (carriers) and uplink is splitted in 24 tones. 2 to 15 bits are coded by tone. The transmission rate is optimized with respect to line conditions. If transmission on one of the tone is disrupted because of strong reflections and interferences at the frequency band, transmission is suspended on that tone by modem.

3.3.1 HDSL

HDSL : High bit rate Digital Subscriber Line is another implementation for symetric speeds (uplink rate = downlink rate). The rate is fixed to 1,6 Mb/s. The sampling rate is 640 KHz, the number of carriers is 256. 256 QAM subsymbols are applied to a treillis encoder and then modulated by a 512 points IFFT.

3.4 Hiperlan/2

3.4.1 Definition

HiperLAN2 is the all new high performance radio technology, specifically suited for operating in LAN environments. HiperLAN2 is a technology being developed within the European Telecommunications Standardisation Institute (ETSI) and a final specification is due to be finalised at the end of 1999 or beginning of 2000.

HiperLAN2 operates in the unlicensed 5 GHz frequency band, which has been specifically allocated to wireless LANs. In contrast to the IEEE 802.11 wireless Ethernet technology, HiperLAN2 is connection-oriented. Connections over the air are time-division multiplexed.

Connections can also be assigned different Quality of Service (QoS). This QoS support allows for the transmission of a mix of different types of technologies, e.g. voice, video, and data. There are also specific connections for unicast, multicast, and broadcast transmission.

HiperLAN2 allows for interconnection into virtually any type of fixed network technology. Thus, HiperLAN2 can carry, for example, Ethernet frames, ATM cells, IP packets, etc. A likely first scenario for HiperLAN2 is to use it between a mobile terminal such as a laptop, and an access point.

OFDM is the modulation used in the physical layer of HiperLAN2, with a 64 point Fast Fourier Transform. For the subcarrier modulation we have choice between BPSK, QPSK, and 16-64 QAM; the symbol period used is 3.6µs with a guard interval of 0.8µs (optionally 0.4µs). The demodulation is coherent. OFDM obviously provides intentionally wide frequency band and a potential bit-rate of 54Mbit/s.

3.4.2 What is the mission statement of H2GF?

Drive the adoption of HiperLAN2 as the globally accepted, broadband wireless technology in the 5GHz band, providing connectivity for mobile devices in corporate, public and home environments.

3.4.3 Who are the founding members?

Bosch, Dell, Ericsson, Nokia, Telia, Texas Instruments

3.4.4 When will the H2GF be launched?

The world-wide launch of HiperLAN2 will take place on the 14th September 1999

in Atlanta and simultaneously in London

3.4.5 How does HiperLAN2 work?

HiperLAN2 operates in the dedicated spectrum in the 5 GHz band. It provides connections, which can be assigned a specific QoS, over the air between the terminal and the base station. While on the move, HiperLAN2 automatically performs handover to the nearest base station (called access point in HiperLAN2). HiperLAN2 also has strong security support, including both authentication and

encryption and has a built-in facility for automatic frequency allocation, removing the need for frequency planning.

3.4.6 Examples of use and scenarios:

With HiperLAN2 installed, workers on a construction site can use laptops to collect blue prints, order materials and communicate with experts. By sending short video sequences via the integrated camera to an expert in real time, a problem can be looked at and discussed, using the high quality audio function, then promptly solved. The broadcast function also means that everyone working on site can be contacted with any information - and that creates a more efficient on-site operation. This application looks like DVB-T and uses OFDM by the same way.

HiperLAN2 benefits companies with a flexible workforce. Employees can transfer their laptops computers from one project to another and continue to exchange large amounts of information between project members and the company server. It is also possible to connect several desktop computers and video projectors via HiperLAN2.

Domestic electronics like televisions, cameras, stereo equipment and computers can all be interconnected by HiperLAN2 using small H2 modules which automatically establish connectivity. HiperLAN2 allows multimedia equipment to be intelligently controlled from any computing device in the home without the need for network cables.

HiperLAN2 enables travellers and employees to work whilst on the move, giving them access to the company network, the Internet as well as allowing them to make and receive multimedia calls. Aircraft Engineers can also benefit from the technology. With customised software, they can access information from databases and get in touch with experts on site.

HiperLAN2 benefits both students and lecturers, allowing wireless access to the university intranet. Covering the entire campus, students can access information, such as videotaped lectures and remote supervision transmitted by their lecturer and two-way communication can take place between students and lecturers through laptops.

3.5 Other applications

• Wireless ATM transmission system. For description, please refer to [17].

• Proposed scheme for UMTS air interface for bit rates higher than 384 kbps (by Telia). But the problem of synchronization of all mobiles in the uplink to the base station is still under study.

4 Simulation

Some results ofthe simulation ofan OFDM system.

4.1 Results

The OFDM model used for the implementation is described in 2.4.

The program has been implemented in matlab. The channel simulation program « propag.c » (has been given to us by a person who doesn’t know the authors of the sources) associated with gene_fad.m computes a vector composed y delays and amplitude of each path. The model used is Raleigh fading.

Here are some results with an OFDM transmission on 1536 carriers with differents modulation schemes on each carrier.

4.1.1 OFDM QAM transmission with only additive noise

4.1.2 OFDM QAM transmission with Raleigh channel model

Constellation at Reception

4.1.3 OFDM BPSK transmission

Constellation at Reoeptioi

4.1.4 BPSK transmission on one carrier

Constellat on at Reception

4.1.5 OFDM 4-PAM transmission

Constellation at leoepticn

0.3

0.2

0.1

o_ 0 0

- -0.1

-0.2

-0.3

-0.4

-0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8

lnphase compoient

5 Conclusion

5.1 Future

5.1.1 Multi Carrier CDMA (MC-CDMA):

The project of MC-CDMA addresses the design of improved wireless radio networks. The mobile or indoor radio channel is characterized by multipath reception: the signal offered to the receiver does not contain only a direct line-of-sight radio wave, which causes significant degradation of the network performances. The effects of multipath radio propagation, modulation and coding added to signal processing techniques on the spectrum efficiency and performance of the wireless radio networks are studied, in particular OFDM and related transmission methods. MC-CDMA intends to further develop the insight in the performance of multi-user systems using OFDM. In particular, this concept is combined with CDMA.

MC-CDMA uses Direct Sequence CDMA (DS-CDMA) merely for multiplexing, but chooses the signal waveforms using the OFDM principle. Indeed, MC-CDMA is a form of DS-CDMA, but after spreading, a Fast Fourier Transform is performed; it can be seen also as a form of OFDM, but we first apply an orthogonal matrix operation to the user bit-stream. Therefore, MC-CDMA is sometimes called CDMA-OFDM. The main improvement of MC-CDMA between OFDM is avoidance of bit errors on subcarriers that are in a deep fade. For that, OFDM generally applies coding (It is the variant of OFDM, which we have called COFDM). Hence the number of subcarriers needed is larger than the number of bits or symbols transmitted simultaneously. MC-CDMA replaces this encoder by an N*N matrix operation. The initial results reveal an improved BER.

The applications of CM-CDMA are almost same than OFDM one and try to achieve best performances. For example about Digital Audio Broadcasting (DAB), MC-CDMA is robust against fading caused by natural multipath and, it can also work if signals are received from two different transmitter sites.

5.2 Conclusion

Because of the great work on third generation mobile networks (UMTS,...) nowadays, everybody speak and think CDMA for digital transmission. However, we’ve seen that OFDM is able to handle very high data rate transmission without too much complexity at the receiver. For that, it is a prefered scheme for digital wideband transmission. So, in the near future, as offering digital wideband transmissions is the new challenge for operators, we might expect to see an explosion in the use of communications systems using OFDM modulation:

• In the subscriber line with XDSL techniques. ADSL commercial tests are currently running in France.

•In the broadband multimedia transmissions with DAB and DVB. DAB networks are being deployed around big cities

• In wireless LAN with HiperLAN/2 standard. HiperLAN/2 allow wide area coverage and for that some people think that such a network may be deployed before UMTS.

However some areas in OFDM are always under researches. This the case for synchronization and coding methods to reduce PAPR.

Eldo Mabiala Mathias Coinchon

6 Annexes

6.1 Bibliography:

1. J.H. Stott :The effects of frequency errors in OFDM. BBC R&D Report BBC RD 1995/15.

2. P. Shelswell : The COFDM Modulation System. BBC Report BBC RD 1996/8.

3. J. Stott :The effect of phase noise in COFDM, BBC R&D

4. Explaining some magic of OFDM

5. J. Proakis : Digital Communications, McGrawHill

6. Sari, Karam, Jeanclaude : Transmission techniques for digital terrestrial TV broadcasting; IEEE communications magazine Febr 1995 vol.33 no.2

7. P. Duhamel : Le système de transmission du projet DAB: Porteuses orthogonales (OFDM)

8. Lazlo Hazy: Initial Channel Estimation and Frame Synchronization in OFDM Systems for Frequency Selective Channels. http://www.sce.carleton.ca/~Laszlo.Hazy/OFDM/

9. John M. Cioffi : A multicarrier primer.

http://www-isl.stanford.edu/~cioffi/papers.html

10. E. Lawrey : The suitability of OFDM as a modulation technique for wireless telecommunications, with a CDMA comparison; http://www.eng.jcu.edu.au/eric/thesis/chapter1.htm#Introduction

11. Speth, Fechtel, Fock, Meyr : Optimum receiver design for wireless broad-band systems using OFDM ; IEEE trans. on communications vol 47 no 11 February 99

12. Chow, Tu : A discrete Multitone Transceiver System for HDSL Applications, IEEE Journal on selected areas in communications vol.9 (6) August 1991.

13. Maddocks, Pullen, Green: Digital Audio Broadcasting, Measuring techniques and coverage performance for a medium power VHF single frequency network. BBC R&D Report BBC RD 1995/2.

14. Bonaccorso, Mhirsi : Reducing peak to average power ratio in OFDM systems. Student Eurecom project 1998.

15. Dutta-Roy :A second wind for Wiring (ADSL) ; IEEE Spectrum spetember 1999

16. ADSL Forum on http://www.adsl.com

17. Prasad : Universal Wireless Personal Communications ; Artech House

18. P.Humblet : Telecommunications; Eurecom course notes Spring 1998

19. G. Caire : Wireless Communications; Eurecom course notes 1999

All BBC R&D articles given here are available on their web site: http://www.bbc.co.uk/rd

A list of interesting links concerning OFDM can be found at : http://studwww.eurecom.fr/~coinchon/ofdm.html

6.2 Glossary

ADSL : Asymetric Digital Subscriber Line

BER : Bit Error Rate

DSP: Digital Signal Processor

ICI : InterCarrier Interference

ISI : InterSymbol Interference

LAN: Local Area Network

Multiplex: mix of several sources

OFDM : Orthogonal Frequency multiplexing

QAM : Quadrature Amplitude Modulation

SFN : Single Frequency Network

VHF : Very High Frequency

6.3 Matlab code

The code has been removed as parts of it comes from industrial partners who want to keep it confidential. Sorry for this inconvenience.

SSCH: Slotted Seeded Channel Hopping for Capacity

Improvement in IEEE 802.11 Ad-Hoc Wireless Networks

Paramvir Bahl

Microsoft Research

One Microsoft Way

Redmond, WA 98052

bahl@microsoft.com

Ranveer Chandra

Department of Computer Science

Cornell University

Ithaca, NY 14853

ranveer@cs.cornell.edu

John Dunagan

Microsoft Research

One Microsoft Way

Redmond, WA 98052

jdunagan@microsoft.com

ABSTRACT

Capacity improvement is one of the principal challenges in wireless networking. We present a link-layer protocol called Slotted Seeded Channel Hopping, or SSCH, that increases the capacity of an IEEE 802.11 network by utilizing fre¬quency diversity. SSCH can be implemented in software over an IEEE 802.11-compliant wireless card. Each node using SSCH switches across channels in such a manner that nodes desiring to communicate overlap, while disjoint communica¬tions mostly do not overlap, and hence do not interfere with each other. To achieve this, SSCH uses a novel scheme for distributed rendezvous and synchronization. Simulation re¬sults show that SSCH significantly increases network capac¬ity in several multi-hop and single-hop wireless networking scenarios.

Categories and Subject Descriptors

C.2.2 [Computer-Communication Networks]: Network Protocols

General Terms

Algorithms, Performance

Keywords

ad-hoc wireless networks, channel assignment, frequency di¬versity, pseudo-randomness, scheduling, medium access con¬trol

1. INTRODUCTION

The problem of supporting multiple senders and receivers in wireless networks has received significant attention in the

Ranveer Chandra was supported, in part, by DARPA un¬der AFRL grant RADC F30602-99-1-0532 and by AFOSR under MURI grant F49620-02-1-0233. Additional support was provided by Microsoft Corporation.

past decade. One domain where this communication pattern naturally arises is fixed wireless multi-hop networks, such as community networks [1, 4, 6, 23]. Increasing the capacity of such wireless networks has been the focus of much recent re¬search (e.g., [14,20,27]). A natural approach to increase the network capacity is to use frequency diversity [9, 31]. Com¬modity wireless networking hardware commonly supports a number of orthogonal channels, and distributing the com¬munication across channels permits multiple simultaneous communication flows.

Channelization was added to the IEEE 802.11 standard to increase the capacity of infrastructure networks — neigh¬boring access points are tuned to different channels so traffic to and from these access points does not interfere [9]. Non-infrastructure (i.e., ad-hoc) networks have thus far been un¬able to capture the benefits of channelization. The current practice in ad-hoc networks is for all nodes to use the same channel, irrespective of whether the nodes are within com¬munication range of each other [4, 6].

In this paper, we propose a new protocol, Slotted Seeded Channel Hopping (SSCH), that extends the benefits of chan-nelization to ad-hoc networks. Logically, SSCH operates at the link layer, but it can be implemented in software over an IEEE 802.11-compliant wireless Network Interface Card (NIC). The SSCH layer in a node handles three aspects of channel hopping (i) implementing the node’s channel hop¬ping schedule and scheduling packets within each channel, (ii) transmitting the channel hopping schedule to neighbor¬ing nodes, and (iii) updating the node’s channel hopping schedule to adapt to changing traffic patterns. SSCH is a distributed protocol for coordinating channel switching de¬cisions, but one that only sends a single type of message, a broadcast packet containing that node’s current channel hopping schedule. Our simulation results show that SSCH yields a significant capacity improvement in ad-hoc wireless networks, including both single-hop and multi-hop scenar¬ios.

The primary research contributions of our paper can be summarized as follows:

Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.

• We present a new protocol that increases the capac¬ity of IEEE 802.11 ad-hoc networks by exploiting fre-quency diversity. This extends the benefits of channel-ization to ad-hoc networks. The protocol is suitable for a multi-hop environment, does not require changes to the IEEE 802.11 standard, and does not require mul¬tiple radios.

• We introduce a novel technique, optimistic synchro-nization, for distributed rendezvous and synchroniza¬tion. This technique allows control traffic to be dis¬tributed across all channels, and thus avoids control channel saturation, a bottleneck identified in prior work on exploiting frequency diversity [31].

• We introduce a second novel technique to achieve good performance for multi-hop communication flows. The partial synchronization technique allows a forwarding node to partially synchronize with a source node and partially synchronize with a destination node. This synchronization pattern allows the load for a single multi-hop flow to be distributed across multiple chan¬nels.

The rest of this paper is organized as follows: we pro¬vide background and motivate the problem in Section 2. In Section 3 we describe SSCH in detail, and in Section 4 we analyze its performance. We discuss design alternatives in Section 5, and we consider related work in Section 6. Fi¬nally, we discuss future work in Section 7, and conclude in Section 8.

2. BACKGROUND AND MOTIVATION

In this paper, we will limit our discussion to the widely-deployed IEEE 802.11 Distributed Coordination Function (DCF) protocol [8]. We begin by reviewing some relevant details of this protocol. IEEE 802.11 recommends the use of a Request To Send (RTS) and Clear To Send (CTS) mech¬anism to control access to the medium. A sender desiring to transmit a packet must first sense the medium free for a DCF interframe space (DIFS). The sender then broadcasts an RTS packet seeking to reserve the medium. If the in¬tended receiver hears the RTS packet, the receiver sends a CTS packet. The CTS reserves the medium in the neighbor¬hood of the receiver, and neighbors do not attempt to send a packet for the duration of the reservation. In the event of a collision or failed RTS, the node performs an exponential backoff. For additional details, we refer the reader to [8].

The IEEE 802.11 standard divides the available frequency into orthogonal (non-overlapping) channels. IEEE 802.11b has 11 channels in the 2.4 GHz spectrum, 3 of which are or¬thogonal, and IEEE 802.11a has 13 orthogonal channels in the 5 GHz spectrum. Packet transmissions on these orthog¬onal channels do not interfere if the communicating nodes on them are reasonably separated (at least 12 inches apart for common hardware [9]).

Using only a single channel limits the capacity of a wire¬less network. For example, consider the scenario in Figure 1 where there are 6 nodes within communication range of each other, all the nodes are on the same channel, and 3 of the nodes have packets to send to distinct receivers. Due to in¬terference on the single channel, only one of them, in this case node 3, can be active. In contrast, if all 3 orthogo¬nal channels are used, all the transmissions can take place simultaneously on distinct channels. SSCH captures the ad¬ditional capacity provided by these orthogonal channels.

We imposed three constraints on ourselves in the design of SSCH:

• SSCH should require only a single radio per node. Some of the previous work on exploiting frequency diversity has proposed that each node be equipped

Figure 1: Only one of the three packets can be trans¬mitted when all the nodes are on the same channel.

with multiple radios [9,33]. Multiple radios draw more power, and energy consumption continues to be a sig¬nificant constraint in mobile networking scenarios. By requiring only a single standards-compliant NIC per node, SSCH faces fewer deployability hurdles than schemes with additional hardware requirements.

• SSCH should use an unmodified IEEE 802.11 proto¬col (including RTS/CTS) when not switching chan¬nels. Requiring standards-compliant hardware allows for easier deployment of this technology.

• SSCH should not cause a logical partition, which we de¬fine to occur when two nodes in communication range are unable to communicate. Because SSCH switches each NIC across frequency channels, different NICs may be on different channels most of the time. De¬spite this, any two nodes in communication range will overlap on a channel with moderate frequency (e.g., at least 10 ms out of every half second) and discovery is accomplished during this time. The mathematical properties of SSCH guarantee that this overlap always occurs.

SSCH exploits frequency diversity using an approach we call optimistic synchronization. SSCH is designed to make the common case be that nodes are aware of each other’s channel hopping schedules, yet SSCH allows any node to change its channel hopping schedule at any time. If node

A has traffic to send to another node B, and A knows B’s hopping schedule, A will probably be able to quickly send to

B by changing its own schedule. In the uncommon case that A does not know B’s schedule, or A has out-of-date informa¬tion about B, then the traffic incurs a latency penalty while A discovers B’s new schedule. The SSCH design achieves this good common case behavior when SSCH is used with a workload where traffic patterns change (i.e., new flows are started) less often than hopping schedule updates are propagated. Because hopping schedule update propagation requires only tens of milliseconds, this is a good workload assumption for many wireless networking scenarios. Our performance evaluation in Section 4 gives absolute numbers for these qualitative claims.

SSCH is designed to work in a single-hop or multi-hop environment, and therefore SSCH must support multi-hop flows. We introduce the partial synchronization technique to allow one node B to follow a channel hopping schedule

that overlaps half the time with another node A, and half the time with a third node C; this is necessary for node B to efficiently forward traffic from node A to node C. Although it is trivially possible for node B to have a channel hopping schedule that is an interleaving of A and C’s schedules, this leaves open how B will represent its schedule when a fourth node desires to synchronize with B. The design for channel hopping that we describe in Section 3.3 resolves this issue.

3. SSCH

SSCH switches each radio across multiple channels so that multiple flows within interfering range of each other can si-multaneously occur on orthogonal channels. This results in significantly increased network capacity when the network traffic pattern consists of such flows.

SSCH is a distributed protocol, suitable for deployment in a multi-hop wireless network. It does not require synchro¬nization or leader election. Nodes do attempt to synchro¬nize, but lack of synchronization results in at most a mild reduction in throughput.

SSCH defines a slot to be the time spent on a single chan¬nel. We choose a slot duration of 10 ms to amortize the over¬head of channel switching. At 54 Mbps (the maximum data rate in IEEE 802.11a), 10 ms is equivalent to 35 maximum-length packet transmissions. A longer slot duration would have further decreased the overhead of channel switching, but would have increased the delay that packets encounter during some forwarding operations. The channel schedule is the list of channels that the node plans to switch to in sub¬sequent slots and the time at which it plans to make each switch. Each node maintains a list of the channel schedules for all other nodes it is aware of – this information is al¬lowed to be out-of-date, but the common case will be that it is accurate. The good performance exhibited by SSCH (Section 4) validates this claim.

We develop the SSCH protocol by first describing its as¬sumptions about the underlying hardware and Medium Ac¬cess Control (MAC) protocol (Section 3.1). We then de¬scribe the packet transmission attempts that are made by each node within a slot, and we refer to this as the packet schedule (Section 3.2). We then define the policy for updat¬ing the channel schedule and for propagating the channel schedule to other nodes (Section 3.3). We then describe the mathematical properties that guided SSCH’s design (Sec¬tion 3.4). Finally, we discuss implementation considerations for SSCH (Section 3.5).

3.1 Hardware and MAC Assumptions

We assume that all nodes are using IEEE 802.11a – SSCH could also be used with other MACs in the IEEE 802.11 family, but this evaluation is beyond the scope of our pa¬per. IEEE 802.11a supports 13 orthogonal channels, and we assume no co-channel interference, a reasonable assumption for physically separated nodes [9]. We expect wireless cards to be capable of switching across channels. As we discuss in more detail at the beginning of Section 4, recent work has re¬duced this switching delay to approximately 80 µs ( [3,17]). We assume that each wireless card contains only a single half-duplex single-channel transceiver.

We require that NICs with a buffered packet wait after switching for the maximum length of a packet transmission before attempting to reserve the medium. This prevents hid¬den terminal problems from occurring just after switching.

This requirement about the hardware is not necessary if the NIC packet buffer can be cleared whenever the channel is switched.

3.2 Packet Scheduling

SSCH maintains packets in per-neighbor FIFO queues. Using FIFO queues maintains standard higher-layer assump¬tions about in-order delivery. The per-neighbor FIFO queues are maintained in a priority queue ordered by perceived neighbor reachability. At the beginning of a slot, packet transmissions are attempted in a round-robin manner among all flows. If a packet transmission to a particular neighbor fails, the corresponding flow is reduced in priority until a pe¬riod of time equal to one half of a slot duration has elapsed – this limits the bandwidth wasted on flows targeted at nodes that are currently on a different channel to at most two packets per slot whenever a flow to a reachable node also exists. Packets are only drawn from the flows that have not been reduced in priority unless only reduced priority flows are available.

Because nodes using SSCH will often be on different chan¬nels, broadcast packets transmitted in any one slot are likely to reach only some of the nodes within physical communi¬cation range. The SSCH layer handles this issue through repeated link-layer retransmission of broadcast packets en-queued by higher layers. Although broadcast packets sent this way may reach a different set of nodes than if all nodes had been on the same channel, we have not found this to present a difficulty to protocols employing broadcast packets — in Section 4 we show that as few as 6 transmissions allows DSR (a protocol that relies heavily on broadcasts) to func¬tion well. This behavior is not surprising because broadcast packets are known to be less reliable than unicast packets, and so protocols employing them are already robust to their occasional loss. However, the SSCH retransmission strategy may not be compatible with all uses of broadcast, such as its use to do synchronization [15]. Also, deploying SSCH in an environment with a different number of channels might require the choice of 6 transmissions to be revisited. Finally, although retransmission increases the bandwidth consumed by broadcast packets, SSCH still delivers significant capacity improvement in the traffic scenarios we studied (Section 4).

An SSCH node with a packet to send may discover that a neighbor is not present on a given channel when no CTS is received in response to a transmitted RTS. However, the node may very well be present on another channel, in which case SSCH should still deliver the packet. To handle this, we initially retain the packet in the packet queue. Packets are dropped only when SSCH gives up on all packets to a given destination, and this dropping of an entire flow occurs only when we have failed to transmit a packet to the destina¬tion node for an entire cycle through the channel schedule. We will explain the meaning of a cycle through the channel schedule in Section 3.3, but with our chosen parameter set¬tings the timeout is 530 ms. After a flow has been garbage collected, new packets with the same destination inserted in the queue are assigned to a new flow, and attempted in the normal manner.

This packet scheduling policy is simple to implement, and yields good performance in the common case that node sched¬ules are known, and information about node availability is accurate. A potential drawback is that a node crash (or other failure event) can lead to a number of wasted RTSs

to the failed node. When added across channels, the num¬ber may exceed the IEEE 802.11 recommended limit of 7 retransmission attempts. In Section 4, we quantify the cost of such failures and show that it is small.

3.3 Channel Scheduling

We begin our description of channel scheduling by de¬scribing the data structure used to represent the channel schedule. We then describe the policy nodes use to act on their own channel schedule, the mechanism to communi¬cate channel schedules to other nodes, and finally the policy nodes implement for updating or changing their own channel schedule.

The channel schedule must capture a given node’s plans for channel hopping in the future, and there is obvious over¬head to representing this as a very long list. Instead, we compactly represent the channel schedule as a current chan¬nel and a rule for updating the channel – in particular, as a set of 4 (channel, seed) pairs. Our experimental results show that 4 pairs suffice to give good performance (Section 4). We represent the (channel, seed) pair as (xi, ai). The channel xi is represented as an integer in the range [0, 12] (13 possi¬bilities), and the seed ai is represented as an integer in the range [1, 12]. Each node iterates through all of the channels in the current schedule, switching to the channel designated in the schedule in each new slot. The node then increments each of the channels in its schedule using the seed,

xi  (xi + ai) mod 13

and repeats the process.

We introduce one additional slot to prevent logical parti¬tions. After the node has iterated through every channel on each of its 4 slots, it switches to a parity slot whose chan¬nel assignment is given by xparity = a1. The term parity slot is derived from the analogy to the parity bits appended at the end of a string in some error correcting codes. The mathematical justification for this design is given in Sec¬tion 3.4. We use the term cycle to refer to the 530 ms iter¬ation through all the slots, including the parity slot.

In Figure 2, we illustrate possible channel schedules for two nodes in the case of 2 slots and 3 channels. In the Fig¬ure, node A and node B are synchronized in one of their two slots (they have identical (channel, seed) pairs), and they also overlap during the parity slot. The field of the channel schedule that determines the channel during each slot is shown in bold. Each time a slot reappears, the chan¬nel is updated using the seed. For example, node A’s slot 1 initially has (channel, seed) = (1,2). The next time slot 1 is entered, the channel is updated by adding the seed to it mod 3 (mod 3 because in this example, there are 3 channels). The resulting channel is given by (1 + 2) mod 3 = 0.

Nodes switch from one slot to the next according to a fixed schedule (every 10 ms in our current parameter settings). However, the decision to switch channels may occur while a node is transmitting or receiving a packet. In this case we delay the switch until after the transmission and ACK (or lack thereof) have occurred.

Nodes learn each other’s schedules by periodically broad¬casting their channel schedule and offset within this cycle. We use the IEEE 802.11 Long Control Frame Header format (see Section 3.5) to embed both the schedule and the node’s current offset. The SSCH layer at each node schedules one of these packets for broadcast once per slot.

Figure 2: Channel hopping schedules for two nodes with 3 channels and 2 slots. Node A always overlaps with Node B in slot 1 and the parity slot. The field of the channel schedule that determines the channel during each slot is shown in bold.

Nodes also update their knowledge of other nodes’ sched¬ules by trying to communicate and failing. Whenever a node sends an RTS to another node, and that node fails to re¬spond even though it was believed to be in this slot, the node sending the RTS updates the channel schedule for the other node to reflect that it does not currently know the node’s schedule in this slot.

We now turn to the question of how a given node changes its own schedule. Schedules are updated in two ways: each node attempts to maintain that its slots start and stop at roughly the same time as other nodes, and that its chan¬nel schedule overlaps with nodes for which it has packets to send. We embed the information needed for this syn¬chronization within the Long Control Frame Header as well. Using this information, a simple averaging scheme such as described by Elson et al [15] can be applied to achieve the loose synchronization required for good performance (Sec¬tion 4 shows that a 100 µs skew in clock times leads to less than a 2% decrease in capacity).

At a high level, each node achieves overlap with nodes for which it has traffic straightforwardly, by changing part of its own schedule to match that of the other nodes. However, a number of minor decisions must be made correctly in order to achieve this high level goal.

Nodes recompute their channel schedule right before they enqueue the packet announcing this schedule in the NIC (and so at least once per slot). In a naive approach, this node could examine its packet queue, and select the (chan¬nel, seed) pairs that lead to the best opportunity to send the largest number of packets. However, this ignores the interest this node has in receiving packets, and in avoiding congested channels. An example of the kind of problem that might arise if one ignores the interest in receiving packets is given in Figure 3. Here, A synchronized with B, and then B synchronized with C in such a way that A was no longer synchronized with B. This could have been avoided if B had used its other slot to synchronize with C, as it would have if it considered its interest in receiving packets.

Figure 3: The problem with a naive synchronization scheme. Node A has two slots, with (channel, seed) pairs represented by A1 and A2; nodes B and C are similarly depicted. At time t1, node A synchronizes with node B. Node B synchronizes with node C at time t2, after which A and B are no longer synchro¬nized.

To account for this node’s interest in receiving packets, we maintain per-slot counters for the number of packets re¬ceived during the previous time the slot was active (ignor¬ing broadcast packets). Any slot that received more than 10 packets during the previous iteration through that slot is labeled a receiving slot; if all slots are receiving slots, any one is allowed to be changed. If some slots are receiving slots and some are not, only the (channel, seed) pair on a non-receiving slot is allowed to be changed for the purpose of synchronizing with nodes we want to send to.

To account for channel congestion, we compare the (chan¬nel, seed) pairs of all the nodes that sent us packets in a given slot with the (channel, seed) pairs of all the other nodes in our list of channel schedules. If the number of other nodes synchronized to the same (channel, seed) pair is more than twice as many as this node communicated with in the pre¬vious occurrence of the slot, we attempt to de-synchronize from these other nodes. De-synchronization just involves choosing a new (channel, seed) pair for this slot. In our experiments, this de-synchronization mechanism was both necessary and sufficient to prevent the nodes from all con¬verging to the same set of (channel, seed) pairs.

The final constraints we add moderate the pace of change in schedule information. Each node only considers updating the (channel, seed) pair for the next slot, never for slots further in the future. If the previous set of criteria suggest updating some slot other than the next slot, we delay that decision. Given these constraints, picking the best possible (channel, seed) pair simply requires considering the choice that synchronizes with the set of nodes for which we have the largest number of queued packets. Additionally, the (channel, seed) pair for the first slot is only allowed to be updated during the parity slot – this helps to prevent logical partition, as will be explained in more detail in Section 3.4.

This strategy naturally supports nodes acting as sources, sinks, or forwarders. A source node will find that it can as¬sign all of its slots to support sends. A sink node will find that it rarely changes its slot assignment, and hence nodes sending to it can easily stay synchronized. A forwarding node will find that some of its slots are used primarily for receiving; after re-assigning the channel and seed in a slot to support sending, the slots that did not change are more likely to receive packets, and hence to stabilize on their cur-rent channel and seed as receiving slots for the duration of

the current traffic patterns. Our simulation results (Sec¬tion 4) support this conclusion. We refer to the technique of enabling this synchronization pattern as partial synchro¬nization.

3.4 Mathematical Properties of SSCH

Our discussion of the mathematical properties of SSCH will initially focus on the static case. The behavior of SSCH when channel schedules are not changing assures us that in a steady-state flow setting, nodes will rendezvous appropri¬ately, in a sense that we make precise below. We will then expand our discussion to include the dynamics of channel scheduling in an environment where flows are starting and stopping.

The channel scheduling mechanism has three simultane¬ous design goals: allowing nodes to be synchronized in a slot, infrequent overlap between nodes that do not have data to send to each other, and ensuring that all nodes come into contact occasionally (to avoid a logical partition). To achieve these goals, we rely on a very simple mathematical technique, addition modulo a prime number.

Consider two nodes that want to be synchronized in a given slot. If they have identical (channel, seed) pairs for this slot, then clearly they will remain synchronized in future iterations (using the static assumption). Now consider two nodes that are not synchronized because they have differ¬ent seeds. A simple calculation shows that these two nodes will overlap exactly one out of every 13 iterations in this slot (recall that 13 is the number of channels). This is the behavior we want from these nodes: they overlap regularly enough that they can exchange their channel schedules, but they are mostly on different channels, and so do not interfere with each other’s transmissions.

Now consider the rare case that two nodes share identical seeds in every slot, but different channels accompany each seed – this has at most a 1 in 134  28, 000 chance of oc¬curring for randomly chosen (channel, seed) pairs. In this case, the nodes will march in lock-step through the same set of channels in each slot, never overlapping. This would be problematic, and it is this situation that the parity slot prevents. To justify this claim, we consider two distinct sit¬uations. If both nodes enter their parity slot at the same time, then they overlap there because the parity channel is equal to the seed for the first slot for both nodes. With our chosen parameter settings of 10 ms per slot, 4 slots, and 13 channels, this overlap occurs once every 530 ms and lasts for 10ms. If their parity slots do not occur at the same time, then the first node’s parity slot offers a fixed target for the slot in which the second node is changing channels, and again, the two nodes will overlap. This overlap occurs once every 7 seconds. Although both these cases will be rare, the SSCH time synchronization mechanism allows us to ignore the second case entirely – a relative clock skew of 5 ms or less is sufficient to guarantee that two parity slots overlap in time.

Now considering the dynamic case (and assuming clock synchronization to within 5 ms), we note that nodes are not permitted to change the seed for the first of their four slots except during a parity slot. Therefore they will always overlap in either the first slot or the parity slot, and hence will always be able to exchange channel schedules within a moderate time interval.

The use of addition modulo a prime to construct channel

hopping schedules does not restrict SSCH to scenarios where the number of channels is a prime number. If one desired to use SSCH with a wireless technology where the number of channels is not a prime, one could straightforwardly use a larger prime as the range of xi, and then map down to the actual number of channels using a modulus reduction. Though the mapping would have some bias to certain chan¬nels, the bias could be made arbitrarily small by choosing a sufficiently large prime.

A final point about the use of addition modulo a prime is that SSCH can be modified to require fewer bits to rep¬resent a node’s schedule by reducing the number of choices for a seed. The only penalty to this reduction is increasing the protocol’s reliance on the parity slot for avoiding logical partitions.

3.5 Implementation Considerations

When simulating SSCH in QualNet [5], we made two tech-nical choices that seem to be relatively uncommon based on our reading of the literature. The first technical choice re¬lates to how we added SSCH to an existing system, and the second relates to a little-utilized part of the IEEE 802.11 specification.

In order to implement SSCH, we had to implement new packet queuing and retransmission strategies. To avoid re¬quiring modifications to the hardware (in QualNet, the hard¬ware model) or the network stack, SSCH buffers packets below the network layer, but above the NIC device driver. To maintain control over transmission attempts, we config¬ure the NIC to buffer at most one packet at a time, and to attempt exactly one RTS for each packet before returning to the SSCH layer. By observing NIC-level counters before and after every attempted packet transmission, we are able to determine whether a CTS was heard for the packet, and if so, whether the packet was successfully transmitted and acknowledged. All the necessary parameters to do this are exposed by the hardware model we used in QualNet. This also prevents head-of-line blocking from interfering with our desire to implement the SSCH transmission strategy.

For efficiency reasons, we choose to use the IEEE 802.11 Long Control Frame Header format to broadcast channel schedules and current offsets, rather than using a full broad¬cast data packet. The most common control frames in IEEE 802.11 (RTS, CTS, and ACK) use the alternative short for¬mat. The long format was included in the IEEE 802.11 standard to support inter-operability with legacy 1-Mbps and 2-Mbps DSSS systems [8]. The format contains 6 un¬used bytes; we use 4 to embed the 4 (channel, seed) pairs, and another 2 to embed the offset within the cycle (i.e., how far the node has progressed through the 530 ms cycle).

Lastly, we comment that the beaconing mechanism used in IEEE 802.11 ad-hoc mode for associating with a Basic Service Set (BSS) works unchanged in the presence of SSCH. A newly-arrived node can associate to a BSS as soon as it overlaps in the same channel with any already-arrived node.

4. SYSTEM EVALUATION

We simulate SSCH in QualNet and compare its perfor¬mance with the commonly used single-channel IEEE 802.11a protocol. In Section 4.1, we present microbenchmarks quan¬tifying the different SSCH overheads. In Section 4.2, we present macrobenchmarks on the performance of SSCH with a large number of nodes in a single hop environment. In Sec¬

tion 4.3, we extend the macrobenchmark evaluation to en¬compass mobility and multi-hop routing. Our results show that SSCH incurs very low overhead, and significantly out¬performs IEEE 802.11a in a multiple flow environment.

Our simulation environment comprises a varying number of nodes in a 200m × 200m area. All nodes in a single simula¬tion run use the same MAC, either SSCH or IEEE 802.11a. We set all nodes to operate at the same raw data rate, 54 Mbps. We assume 13 usable channels in the 5 GHz band. SSCH is configured to use 4 seeds, and each slot duration is 10 ms. All seeds are randomly chosen at the beginning of each simulation run. The macrobenchmarks in Sections 4.2 and 4.3 are averages from 5 independent simulation runs, while the microbenchmarks in Section 4.1 are drawn from a single simulation run.

We primarily measure throughput under a traffic load of maximum rate UDP flows. In particular, we use Constant Bit Rate (CBR) flows of 512 byte packets sent every 50 µs. This data rate is more than the sustainable throughput of IEEE 802.11a operating at 54 Mbps.

For all our simulations, we modified QualNet to use a channel switch delay of 80 µs. This choice was based on re¬cent work in solid state electronics on reducing the settling time of the Voltage Control Oscillator (VCO) [7]. Switching the channel of a wireless card requires changing the input voltage of the VCO, which operates in a Phase Locked Loop (PLL) to achieve the desired output frequency. The delay in channel switching is due to this settling time. The speci¬fication of Maxim IEEE 802.11b Transceivers [3] shows this delay to be 150 µs. A more recent work [17] shows that this delay can be reduced to 40-80 µs for IEEE 802.11a cards.

4.1 Microbenchmarks

We present microbenchmarks measuring the overhead of SSCH in several different scenarios. In Section 4.1.1, we measure the overhead during the successful initiation of a CBR flow. In Section 4.1.2, we measure the overhead on an existing session of failing to initiate a parallel CBR flow. In Section 4.1.3, we measure the overhead of supporting two flows simultaneously. In Section 4.1.4, we measure the over-head of continuing to attempt transmissions to a mobile node that has moved out of range. These scenarios cover many of the different dynamic events that a MAC must ap¬propriately handle: a flow starting while a node is present, a flow starting while a node is absent, simultaneous flows where both nodes are present, simultaneous flows where one node moves out of range, etc. Finally, the scenario in Sec¬tion 4.1.5 measures the overhead of SSCH with respect to a different kind of event, clock skew.

4.1.1 Overhead of Switching and Synchronizing

In this experiment, we measured the overhead of success¬fully initiating a CBR flow between two nodes within com¬munication range of each other. The first node initiates the flow just after the parity slot. This incurs a worst-case delay in synchronization, because the first of the four slots will not be synchronized until 530 ms later.

Figure 4 shows the moving average over 20 ms of the throughput at the receiver node. The sender quickly syn¬chronizes with the receiver on three of the four slots, as it should, and on the fourth slot after 530 ms. The fig¬ure shows the throughput while synchronizing (oscillating around 3/4 of the raw bandwidth), and the time required

10 12 14 16 18 20 22 24

Time (in seconds)

SSCH 802.11a

Figure 4: Switching and Synchronizing Overhead: Node 1 starts a maximum rate UDP flow to Node 2. We show the throughput for both SSCH and IEEE 802.11a.

to synchronize. After synchronizing, the channel switch¬ing and other protocol overheads of SSCH lead to only a 400 Kbps penalty in the steady-state throughput relative to IEEE 802.11a. This penalty conforms to our intuition about the overheads in SSCH: a node spends 80 µs every 10 ms switching channels (80 µs/10 ms = .008), and then must wait for the duration of a single packet to avoid colliding with pre-existing packet transmissions in the new channel (1 packet/35 packets = .028). Adding these two overheads together leads to an expected cumulative overhead of 3.6%, which is in close agreement with the measured overhead of (400 Kbps/12 Mbps) = 3.3%.

Note that the throughput reaches a maximum of only 13 Mbps, although the raw data rate is 54 Mbps. This low uti¬lization can be explained by the IEEE 802.11a requirement that the RTS/CTS packets be sent at the lowest supported data rate, 6 Mbps, along with other overheads [18].

4.1.2 Overhead of an Absent Node

SSCH requires more re-transmissions than IEEE 802.11 to prevent logical partitions. These retransmissions waste bandwidth that could have been dedicated to a node that was present on the channel. To quantify this overhead, we initiated a CBR flow between two nodes, allowed the system to quiesce, and then initiated a send from the first node to a non-existent node. Figure 5 shows the moving average over 80 ms of the throughput. The Figure shows that the sender takes 530 ms to timeout on the non-existent node. During this time the throughput drops by 550 Kbps, which is a small fraction (4.6%) of the total throughput.

4.1.3 Overhead of a Parallel Session

Our next experiment quantifies the ability of SSCH to fairly share bandwidth between two flows, and to quickly achieve this fair sharing. To measure this we start with Node 1 sending a maximum rate UDP flow to Node 2. At 21.5 seconds, Node 1 starts a second maximum rate UDP flow to Node 3. Figure 6 presents the moving average over

23 23.5 24 24.5 25 25.5 26 26.5 27

Time (in seconds)

Figure 5: Overhead of an Absent Node: Node 1 is sending a maximum rate UDP flow to Node 2. Node 1 then attempts to send a packet to a non-existent node.

15 20 25 30 35

Time (in seconds)

Node 2 Node 3

Figure 6: Overhead of a Parallel Session: Node 1 is sending a maximum rate UDP flow to Node 2. Node 1 then starts a second flow to Node 3.

140 ms of the throughput achieved by both receivers. The bandwidth is split between the receivers nearly perfectly, and with no decrease in net throughput.

4.1.4 Overhead of Mobility

We now analyze the effect of mobility at a micro-level on the performance of SSCH. Ideally, SSCH should be able to detect a link breakage due to movement of a node, and sub¬sequently re-synchronize to other neighbors. We show that SSCH can indeed handle this scenario with an experiment comprising 3 nodes and 2 flows, and in Figure 7 we present the moving average over 280 ms of each flow’s throughput.

Node 1 is initially sending a maximum rate UDP flow to Node 2. Node 1 initiates a second UDP flow to Node 3 at around 20.5 seconds. This bandwidth is then shared be¬tween both the flows (as in the experiment of Section 4.1.3) until 30 seconds, when Node 3 moves out of the communi¬cation range of Node 1. Our experiment configures Node 1 to continue to attempt to send to Node 3 until 43 seconds,

Node 2 Node 3

Figure 7: Overhead of Mobility: Node 1 is sending a maximum rate UDP flow to Node 2. Node 1 starts another maximum rate UDP flow to Node 3. Node 3 moves out of range at 30 seconds, while Node 1 continues to attempt to send until 43 seconds.

1 ns 10 ns 100 ns 1 µs 10 µs 100 µs 1 ms

Clock Drift

Figure 8: Overhead of Clock Skew: Throughput be¬tween two nodes using SSCH as a function of clock skew.

and during this time it continues to consume a small amount of bandwidth. In contrast, the experiment in Section 4.1.2 measured the overhead of enqueueing a single packet to an absent node. When the flow to Node 3 finally stops, Node 2’s received throughput increases back to its initial rate.

4.1.5 Overhead of Clock Drift

As we described in Section 3.3, SSCH tries to synchronize slot begin and end times, though it is also designed to be robust to clock skew. In this experiment, we quantify the robustness of SSCH to moderate clock skew. We measure the throughput between two nodes after artificially intro¬ducing a clock skew between them, and disabling the SSCH synchronization scheme for slot begin and end times. We vary the clock skew from 1 ns (106 ms) to 1 ms such that the sender is always ahead of the receiver by this value, and present the results in Figure 8. Note the log scale on the x-axis.

The throughput achieved between the two nodes is not significantly affected by a clock skew of less than 10 µs.

These practical values of clock skew are extremely small to impact the performance of SSCH. The drop in throughput is more for larger clock skews, although the throughput is still acceptable at 10.5 Mbps when the skew value is an extremely high 1 ms.

These results provide justification for the design choice we made of not requiring nodes to switch synchronously across slots, as described in Section 3.3. For example, a node will delay switching to receive an ACK, or to send a data packet if its channel reservation is successful. In the 100 node ex-periment described in Section 4.3.2, we measured the skew in channel switching times for a traffic pattern of 50 flows to be approximately 20 µs. Figure 8 shows that this is a negligible amount.

4.2 Macrobenchmarks: Single-hop Case

We now present simulation results showing SSCH’s ability to achieve and sustain a consistently high throughput for a traffic pattern consisting of multiple flows. We first evaluate this using steady state UDP flows. We then extend our evaluation to consider a dynamic traffic scenario where UDP flows both start and stop. Finally, we study the performance of TCP over SSCH.

4.2.1 Disjoint Flows

We first look at the number of disjoint flows that can be supported by SSCH. All nodes in this experiment are in communication range of each other, and therefore two flows are considered disjoint if they do not share either endpoint. Ideally, SSCH should utilize the available bandwidth on all the channels on increasing the number of disjoint flows in the system. We evaluate this by varying the number of nodes in the network from 2 to 30 and introducing a flow between disjoint pairs of nodes — the number of flows varies from 1 to 15.

Figure 9 shows the per-flow throughput, and Figure 10 shows the total system throughput. IEEE 802.11a performs marginally better when there is just one flow in the network. When there is more than one flow, SSCH significantly out¬performs IEEE 802.11a.

An increase in the number of flows decreases the per-flow throughput for both SSCH and IEEE 802.11a. However, the drop for IEEE 802.11a is much more significant. The drop for IEEE 802.11a is easily explained by Figure 10, which shows that the overall system throughput for IEEE 802.11a is approximately constant.

It may seem surprising that the SSCH system throughput has not stabilized at 13 times the throughput of a single flow by the time there are 13 flows. However, this can be attributed to SSCH’s use of randomness to distribute flows across channels. These random choices do not lead to a perfectly balanced allocation, and therefore there is still un¬used spectrum even when there are 13 flows in the system, as shown by the continuing positive slope of the curve in Figure 9.

4.2.2 Non-disjoint Flows

We now consider the case when the flows in the network are not disjoint – nodes participate as both sources and sinks, and in multiple flows. This scenario stresses SSCH’s ability to efficiently support sharing among simultaneous flows that have a common endpoint. Each node in the net¬work starts a maximum rate UDP flow with one other ran

2 4 6 8 10 12 14 16

# Nodes

Figure 13: Effect of Flow Duration: Ratio of SSCH throughput to IEEE 802.11a throughput for flows having different durations.

2 3 4 5 6 7 8

# Flows

Figure 14: TCP over SSCH: Steady-state TCP throughput when varying the number of non-disjoint flows.

evaluate this concern quantitatively, we run an experiment where we vary the number of nodes in the network from 2 to 9, such that all nodes are in communication range of one another. We then start an infinite-size file transfer over FTP from each node to a randomly selected other node. This choice to use non-disjoint flows is designed to stress the SSCH implementation by requiring nodes to be synchronized as either senders or receivers with multiple other nodes. In Figure 14 we present the resulting cumulative steady-state TCP throughput over all the flows in the network.

Figure 14 shows that the TCP throughput for a small number of flows is lower for SSCH than the throughput over IEEE 802.11a. However, as the number of flows increases, SSCH does achieve a higher system throughput. Although TCP over SSCH does provide higher aggregate throughput than over IEEE 802.11a, the performance improvement is not nearly as good as for UDP flows. This shows that jitter due to SSCH does have an impact on the performance of

0 6 12 18

# Nodes

SSCH 802.11a

Figure 15: Multi-hop Chain Network: Variation in throughput as chain length increases.

TCP. A more detailed analysis of the interaction between TCP and SSCH, and modifications to support better inter¬actions between TCP and SSCH, is a subject we plan to address in our future work.

4.3 Macrobenchmarks: Multi-hop Case and Mobility

We now evaluate SSCH’s performance when combined with multi-hop flows and mobile nodes. We first analyze the behavior of SSCH in a multi-hop chain network. We then consider large scale multi-hop networks, both with and without mobility. As part of this analysis, we study the interaction between SSCH and MANET routing protocols.

4.3.1 Performance in a Multi-hop Chain Network

IEEE 802.11 is known to encounter significant perfor¬mance problems in a multi-hop network [34]. For example, if all nodes are on the same channel, the RTS/CTS mechanism allows at most one hop in an A  B  C  D chain to be ac¬tive at any given time. SSCH reduces the throughput drop due to this behavior by allowing nodes to communicate on different channels. To examine this, we evaluate both SSCH and IEEE 802.11a in a multi-hop chain network.

We vary the number of nodes, which are all in communi-cation range, from 2 to 18. We initiate a single flow that encounters every node in the network. Although more than 4 nodes transmitting within interference range of each other would be unlikely to arise from multi-hop routing of a single flow, it could easily arise in a general distributed application. Figure 15 shows the maximum throughput as the number of nodes in the chain is varied. We see that there is not much difference between SSCH and IEEE 802.11a for flows with few hops. As the number of hops increases, SSCH performs much better than IEEE 802.11a since it distributes the com-munication on each hop across all the available channels.

4.3.2 Performance in a Multi-hop Mesh Network

We now analyze the performance of SSCH in a large scale multi-hop network without mobility. We place 100 nodes uniformly in a 200 × 200 m area, and set each node to trans¬mit with a power of 21 dBm. The Dynamic Source Routing

Figure 16: Mulithop Mesh Network of 100 Nodes: The per-flow throughput on varying the number of flows in the network.

(DSR) [22] protocol is used to discover the source route be¬tween different source-destination pairs. These source routes are then input to a static variant of DSR that does not per¬form discovery or maintain routes. We vary the number of maximum rate UDP flows from 10 to 50. We generate source-destination pairs by choosing randomly, and reject¬ing pairs that are within a single hop of each other.

We present the per-flow throughput in Figure 16. Increas¬ing the number of flows leads to greater contention, and the average throughput of both SSCH and IEEE 802.11a drops. For every considered number of flows, SSCH provides signif¬icantly higher throughput than IEEE 802.11a. For 50 flows, the inefficiencies of sharing a single channel are sufficiently pronounced that SSCH yields more than a factor of 15 ca¬pacity improvement.

4.3.3 Impact of Channel Switching on

MANET Routing Protocols

Previous work on multi-channel MACs has often over¬looked the effect of channel switching on routing protocols. Most of the proposed protocols for MANETs rely heavily on broadcasts (e.g., DSR [22] and AODV [28]). However, neighbors using a multi-channel MAC could be on different channels, which could cause broadcasts to reach significantly fewer neighbors than in a single-channel MAC. SSCH ad¬dresses this concern using a broadcast retransmission strat¬egy discussed in Section 3.2.

We study the behavior of DSR [22] over SSCH in the same experimental setup used in Section 4.3.2, with 100 nodes in a 200 m×200 m area. However, we reduce the transmission power of each node to 16 dBm to force routes to increase in length (and hence to stress DSR over SSCH). We select 10 source-destination pairs at random, and we use DSR to discover routes between them. In Figure 17 we compare the performance of DSR over SSCH, when varying the SSCH broadcast transmission count parameter (the number of con¬secutive slots in which each broadcast packet is sent once).

Figure 17 shows that the performance of DSR over SSCH improves with an increase in the broadcast transmission count. The DSR Route Request packets see more neighbors when SSCH broadcasts them over a greater number of slots.

Figure 17: Impact of SSCH on Unmodified MANET Routing Protocols: The average time to discover a route and the average route length for 10 randomly chosen routes in a 100 node network using DSR over SSCH.

This increases the likelihood of discovering shorter routes, and the speed with which routes are discovered. However, there seems to be little additional benefit to increasing the broadcast parameter to a value greater than 6. We attribute the slight bumpiness in the curves to the stochastic nature of DSR, and its reliance on broadcasts.

Comparing SSCH to IEEE 802.11a, we see that the SSCH discovers routes that are comparable in length. However, the average route discovery time for SSCH is much higher than for IEEE 802.11a. Because each slot is 10 ms in length, broadcasts are only retransmitted once every 10 ms, and this leads to a significantly longer time to discover a route to a given destination node. We believe that this latency is a fundamental difficulty in using a reactive protocol such as DSR with SSCH. We plan to explore the interaction of other proactive and hybrid routing protocols with SSCH in the future.

4.3.4 Performance in Multi-hop Mobile Networks

We now study the impact of mobility in a network using DSR over IEEE 802.11a and SSCH. In this experiment, we place 100 nodes randomly in a square and select 10 flows. Each node transmits packets at 21 dBm. Node movement is determined using the Random Waypoint model. In this model, each node has a predefined minimum and maximum speed. Nodes select a random point in the simulation area, and move towards it with a speed chosen randomly from the interval. After reaching its destination, a node rests for a period chosen from a uniform distribution between 0 and 10 seconds. It then chooses a new destination and repeats the procedure. In our experiments, we fix the minimum speed at 0.01 m/s and vary the maximum speed from 0.2 to 1.0 m/s. Although we have studied SSCH at higher speeds, the results are not significantly different. We performed this experiment using two different areas for the nodes, a 200m × 200m area and a 300m × 300m area. We refer to the smaller area as the dense network, and the larger area as the sparse network – the average path is 0.5 hops longer

Figure 18: Dense Multi-hop Mobile Network: The per-flow throughput and the average route length for 10 flows in a 100 node network in a 200m × 200m area, using DSR over both SSCH and IEEE 802.11a.

Figure 19: Sparse Multi-hop Mobile Network: The per-flow throughput and the average route length for 10 flows in a 100 node network in a 300m × 300m area, using DSR over both SSCH and IEEE 802.11a.

in the sparse network. For all these experiments, we set the SSCH broadcast transmission count parameter to 6.

Figure 18 shows that in a dense network, SSCH yields much greater throughput than IEEE 802.11a even when there is mobility. Although DSR discovers shorter routes over IEEE 802.11a, the ability of SSCH to distribute traffic on a greater number of channels leads to much higher overall throughput. Figure 19 evaluates the same benchmarks in a sparse network. The results show that the per-flow through¬put decreases in a sparse network for both SSCH and IEEE 802.11a. This is because the route lengths are greater, and it takes more time to repair routes. However, the same qual¬itative comparison continues to hold: SSCH causes DSR to discover longer routes, but still leads to an overall capacity improvement.

DSR discovers longer routes over SSCH than over IEEE 802.11a because broadcast packets sent over SSCH may not reach a node’s entire neighbor set. Furthermore, some opti¬mizations of DSR, such as promiscuous mode operation of nodes, are not as effective in a multi-channel MAC such as SSCH. Thus, although the throughput of mobile nodes us¬ing DSR over SSCH is much better than their throughput over IEEE 802.11a, we conclude that a routing protocol that takes the channel switching behavior of SSCH into account will likely lead to even better performance.

5. DISCUSSION

In this Section we discuss alternative designs for SSCH within the constraints that we enumerated in Section 2. We will discuss prior work related to SSCH in detail in Section 6.

SSCH distributes the rendezvous and control traffic across all the channels. One straightforward alternative scheme, which still only requires one radio, is to use one of the chan¬nels as a control channel, and all the other channels as data channels (e.g., [21]). Each node must then somehow split its time between the control channel and the data channels.

Such a scheme will have difficulty in preventing the con¬trol channel from becoming a bottleneck. Suppose that two nodes exchange RTS/CTS on the control channel, and then

switch to a data channel to do transmission. Unless all other nodes were also on the control channel during the RTS/CTS exchange, these two nodes will still need to do an RTS/CTS on this channel in order to avoid the hidden terminal prob¬lem. The two nodes should wait to even do the RTS/CTS until after an entire packet transmission interval has elapsed, because another pair of nodes might have also switched to this channel, orchestrating that decision on the control chan¬nel during a time that the first pair of nodes were not on the control channel. In order to amortize this startup cost, the nodes should have several packets to send to each other. However, while any one node remains on a data channel, any other node that desires to send it a packet must remain idle on the control channel waiting for the node it desires to reach to re-appear. If the idle node on the control channel chooses not to wait, and instead switches to a data channel with another node for which it has traffic, it may repeatedly fail to rendezvous with the first node, leading to a significant imbalance in throughput and possibly a logical partition.

The problems with a dedicated control channel may be solvable, but it is clear that a straightforward approach with un-synchronized rendezvous presents several difficulties. If one instead tried to synchronize rendezvous on the control channel, the control channel could again become a bottle¬neck simply because many nodes simultaneously desire to schedule packets on that channel.

6. RELATED WORK

We divide the prior work relevant to SSCH into three categories: prior uses of pseudo-random number generators in wireless networking, channel switching for reasons other than capacity improvement, and alternative approaches to exploiting frequency diversity. In the first category, we find that pseudo-random number generators have been used for a variety of tasks in wireless networking. For example, the SEEDEX protocol [29] uses pseudo-random generators to avoid RTS/CTS exchanges in a wireless network. Nodes build a schedule for sending and listening on a network, and publish their seeds to all the neighbors. A node will attempt

a transmission only when all its neighbors (including the receiver) are in a listening state. Assuming relatively con¬stant wireless transmission ranges, this protocol also helps in overcoming the hidden and exposed terminal problem. The TSMA protocol [11,12] is a channel access scheme pro¬posed as an alternative to ALOHA and TDMA, for time-slotted multihop wireless networks. TSMA aims to achieve the guarantees of TDMA without incurring the overhead of transmitting large schedules in a mobile environment. Each node is bootstrapped with a fixed seed that determines its transmission schedule. The schedules are constructed using polynomials over Galois fields (which have pseudo-random properties), and the construction guarantees that each node will overlap with only a single other node within a certain time frame. The length of the schedule depends on the num-ber of nodes and the degree of the network. Porting these schedules to a multichannel scenario, where the number of channels is fixed, remains an open problem, and even such a porting would not meet the SSCH goal of supporting traffic-driven overlap. Redi et al. [13] use a pseudo-random gen¬erator to derive listening schedules for battery-constrained devices. Each device’s seed is known to a base station, which can then schedule transmissions for the infrequent moments when the battery-constrained device is awake. Although pseudo-random generators have been used for a number of tasks (as this survey of the literature makes clear), to the best of our knowledge, SSCH is the first protocol to use a pseudo-random generator to construct a channel hopping schedule.

The second category of prior work is channel switching for reasons other than capacity improvement. MultiNet [10] is the main piece of work that we are aware of in this cate¬gory. MultiNet allows a NIC to periodically hop between two channels, enabling a single wireless NIC to connect to two logically distinct networks, such as an AP network and an ad-hoc network. MultiNet is designed to provide new functionality: simultaneous connectivity to distinct net¬works using a single NIC. In contrast, SSCH is designed to yield capacity improvement within a single ad-hoc network.

The third category of prior work we define encompasses all prior approaches to increasing network capacity by ex¬ploiting frequency diversity. This is a significant body of work. The first division we make in this body of work is between research that assumes a single NIC capable of com¬municating on a single channel at any given instance in time, and research that assumes more powerful radio technology, such as multiple NICs [9,30] or NICs capable of listening on many channels simultaneously [21, 26], even if they can only communicate on one. Our work falls in to the former cate¬gory; the SSCH architecture can be deployed over a single standards-compliant NIC supporting fast channel switching.

Dynamic Channel Assignment (DCA) [33] and Multi-radio Unification Protocol (MUP) [9] are both technologies that use multiple radios (in both cases, two radios) to take advan¬tage of multiple orthogonal channels. DCA uses one radio on a control channel, and the other radio switches across all the other channels sending data. Arbitration for channels is embedded in the RTS and CTS messages, and is exe-cuted on the control channel. Although this scheme may fully utilize the data channel, it does so at the cost of us¬ing an entire radio just for control. MUP uses both radios for data and control transmissions. Radios are assigned to orthogonal channels, and a packet is sent on the radio with

better channel characteristics. This scheme gives good per-formance in many scenarios. However, it still only allows the use of as many channels as there are radios on each physi¬cal node. From our perspective, the key drawback to both DCA and MUP is simply that they require the use of mul¬tiple radios. Recently, commercial products have appeared that claim the ability to place multiple radios on a single NIC [2]. It is still not known whether these products will ever achieve as many radios on a NIC as there are available channels, nor what their power consumption will be.

A straightforward way to view the different potential gains of SSCH compared to a true multiple radio design is to con¬sider two distinct sources of bottleneck in a single-radio, single-channel system: the saturation of the channel, and the saturation of any particular radio. Conceptually, SSCH significantly increases the channel bandwidth, without in¬creasing the bandwidth of any individual radio. In contrast, a true multiple radio design increases both. A specific ex¬ample of this difference is that a node using MUP (a true multiple radio design) can simultaneously send and receive packets on separate channels, while a node using SSCH can only perform one of these operations at a time.

We next turn our attention to work assuming more pow¬erful radio technology than is currently technologically fea¬sible. HRMA [35] is designed for frequency hopping spread spectrum (FHSS) wireless cards. Time is divided into slots, each one of which corresponds to a small fraction of the time required to send a packet, and the wireless NIC is on a dif¬ferent frequency during each slot. All nodes are required to maintain synchronized clocks, where the synchronization is at the granularity of slot times that are much shorter than the duration of a packet. Each slot is subdivided in to four segments of time for four different possible communica¬tions: HOP-RESERVED/RTS/CTS/DATA. The first three segments of time are assumed to be small in comparison with the amount of time spent sending a segment of the packet during the DATA time interval. To the best of our knowl¬edge, a FHSS wireless card that supports this type of MAC protocol at high data rates is not commercially available.

Another line of related work assumes technology by which nodes can concurrently listen on all channels. For example, Nasipuri et al [26] and Jain et al [21] assume wireless NICs that can receive packets on all channels simultaneously, and where the channel for transmission can be chosen arbitrarily. In these schemes, nodes maintain a list of free channels, and either the sending or receiving node chooses a channel with the least interference for its data transfer. Wireless NICs do not currently support listening on arbitrarily many channels, and we do not assume the availability of such technology in the design of SSCH.

We finally consider prior work that only assumes the pres¬ence of a single NIC with a single half-duplex transceiver. The only other approach that we are aware of to exploit¬ing frequency diversity under this assumption is Multichan¬nel MAC (MMAC) [31]. Like SSCH, MMAC attempts to improve capacity by arranging for nodes to simultaneously communicate on orthogonal channels. Briefly, MMAC oper¬ates as follows: nodes using MMAC periodically switch to a common control channel, negotiate their channel selections, and then switch to the negotiated channel, where they con¬tend for the channel as in IEEE 802.11. This scheme raises several concerns that SSCH attempts to overcome. First, MMAC extends IEEE 802.11 Power Save Mode (PSM) for

ad-hoc networks. This implies a relatively stringent reliance on clock synchronization, which is particularly hard to pro¬vide in multi-hop wireless networks [19]. In contrast, SSCH does not require tight clock synchronization because SSCH does not have a common control channel or a dedicated neighbor discovery interval. Secondly, synchronization traf¬fic in MMAC can be a significant fraction of the system traf¬fic, and the common synchronization channel can become a bottleneck on system throughput. SSCH addresses this concern by distributing synchronization and control traf¬fic across all the available channels. A third concern with MMAC is that it can lead to packet delays of hundreds of milliseconds for even a single hop. MMAC switches channels every 100 ms, so a node with packets for two different desti¬nations will have to wait at least 100 ms to send traffic to one of them whenever the two destinations decide to use different channels. In contrast, SSCH performs well while switching channels every 10 ms. A fourth concern with MMAC is that it does not specify how to support broadcasts, which are required by most MANET routing protocols (e.g., DSR). SSCH addresses this using a broadcast retransmission strat¬egy that we experimentally validated to be compatible with DSR.

Although this survey does not cover all related work, it does characterize the current state of the field. At the level of detail in this section, prior work such as CHMA [32] is similar to HRMA [35], and MAC-SCC [25] and the MAC protocols implicit in the work of Li et al [24] and Fitzek et al [16] are similar to DCA [33]. However, a final related channel hopping technology that is worth mentioning is the definition of FHSS channels in the IEEE 802.11 [8] specifica¬tion. At first glance, it may seem redundant that SSCH does channel hopping across logical channels, each one of which (per the IEEE 802.11 specification) may be employing fre¬quency hopping across distinct frequencies at the physical later. The IEEE 802.11 specification justifies this physi¬cal layer frequency hopping with the scenario of providing support for multiple Basic Service Sets (BSS’s) that can co¬incide geographically without coinciding on the same logical channel. In contrast, SSCH does channel hopping so that any two nodes can coincide as much or as little of the time as they desire. This is also at the heart of the difference between SSCH and past work on channel-hopping proto¬cols where nodes overlap a fixed fraction of the time [12] – the degree of overlap between any two nodes using SSCH is traffic-dependent.

7. FUTURE RESEARCH

SSCH is a promising technology. In our future work, we plan to investigate how SSCH will perform when imple¬mented over actual hardware, and is subject to the normal environmental vagaries of wireless networks, such as unpre¬dictable variations in signal strength. As part of this im¬plementation effort, we also plan to evaluate how metrics reflecting environmental conditions, such as ETX [14], can be integrated into SSCH.

Our results in Section 4.3.3 show that running existing routing protocols over SSCH can incur a large route discov¬ery time and an increase in the route length. In the future, we plan to more thoroughly evaluate routing over SSCH (as opposed to classical single channel routing), and to explore SSCH’s interaction with a wider variety of proactive and hybrid routing protocols.

There are at least four additional topics that would also need to be addressed before SSCH can be deployed. One is interoperability with nodes that are not running SSCH. Another is the evaluation of power consumption under this scheme. We have not attempted to evaluate the energy cost of switching channels, nor have we attempted to enable a power-saving strategy such as in the IEEE 802.11 specifica-tion for access-point mode. A third topic of investigation is the evaluation of SSCH in conjunction with auto-rate adap¬tation mechanisms. A fourth topic is a more detailed eval¬uation of the interplay between SSCH and TCP.

8. CONCLUSION

We have presented SSCH, a new protocol that extends the benefits of channelization to ad-hoc networks. This protocol is compatible with the IEEE 802.11 standard, and is suitable for a multi-hop environment. SSCH achieves these gains using a novel approach called optimistic synchronization. We expect this approach to be useful in additional settings beyond channel hopping.

We have shown through extensive simulation that SSCH yields significant capacity improvement in a variety of single-hop and multi-hop wireless scenarios. In the future, we look forward to exploring SSCH in more detail using an imple-mentation over actual hardware.

Acknowledgment

The authors would like to thank Ken Birman for his insight¬ful comments on early drafts of this paper.

9. REFERENCES

[1] Bay Area Wireless Users Group, http://www.bawug.org.

[2] Engim, http://www.engim.com/.

[3] Maxim 2.4GHz 802.11b Zero-IF Transceivers. http://pdfserv.maxim-ic.com/en/ds/MAX2820-MAX2821.pdf.

[4] MIT RoofNet, http://www.pdos.lcs.mit.edu/roofnet/.

[5] QualNet, http://www.qualnet.com/.

[6] Seattle Wireless, http://www.seattlewireless.net/.

[7] Tracking Advances in VCO Technology, http://pdfserv.maxim-ic.com/en/an/AN1768.pdf.

[8] IEEE 802.11b/D3.0, Wireless LAN Medium Access Control(MAC) and Physical (PHY) Layer Specification: High Speed Physical Layer Extensions in the 2.4 GHz Band, 1999.

[9] A. Adya, P. Bahl, J. Padhye, A. Wolman, and L. Zhou. A Multi-Radio Unification Protocol for IEEE 802.11 Wireless Networks. In IEEE International Conference on Broadband Networks (Broadnets) 2004.

[10] R. Chandra, P. Bahl, and P. Bahl. MultiNet: Connecting to Multiple IEEE 802.11 Networks Using a Single Wireless Card. In IEEE Infocom 2004.

[11] I. Chlamtac and A. Farago. Making Transmission Schedules Immune to Topology Changes in Multi-Hop Packet Radio Networks. IEEE/ACM Transactions on Networking, 2(1):23–29, February 1994.

[12] I. Chlamtac and A. Farago. Time-Spread Multiple-Access (TSMA) Protocols for Multihop Mobile Radio Networks. IEEE/ACM Transactions on Networking, 5(6):804–812, December 1997.

[13] I. Chlamtac, C. Petrioli, and J. Redi.

Energy-Conserving Access Protocols for Identification Networks. IEEE/ACM Transactions on Networking, 7(1):51–61, February 1999.

[14] D. D. Couto, D. Aguayo, J. Bicket, and R. Morris. A High-Throughput Path Metric for Multi-Hop Wireless Routing. In ACM MobiCom 2003.

[15] J. Elson, L. Girod, and D. Estrin. Fine-Grained Network Time Synchronization using Reference Broadcast. In USENIX OSDI 2002.

[16] F. Fitzek, D. Angelini, G. Mazzini, and M. Zorzi. Design and performance of an enhanced IEEE 802.11 MAC protocol for multihop coverage extension. IEEE Wireless Communications, 10(6):30–39, December 2003.

[17] F. Herzel, G. Fischer, and H. Gustat. An Integrated CMOS RF Synthesizer for 802.11a Wireless LAN. IEEE Journal of Solid-state Circuits, 18(10), October 2003.

[18] M. Heusse, F. Rousseau, G. Berger-Sabbatel, and

A. Duda. Performance Anomaly of 802.11b. In IEEE

Infocom 2003.

[19] L. Huang and T.-H. Lai. On the scalability of IEEE 802.11 ad hoc networks. In ACM MobiHoc 2002.

[20] K. Jain, J. Padhye, V. Padmanabhan, and L. Qiu. Impact of Interference on Multi-hop Wireless Network Performance. In ACM MobiCom 2003.

[21] N. Jain and S. R. Das. A Multichannel CSMA MAC Protocol with Receiver-Based Channel Selection for Multihop Wireless Networks. In IEEE International Conference on Computer Communications and Networks (IC3N) 2001.

[22] D. Johnson, D. Maltz, and J. Broch. DSR: The Dynamic Source Routing Protocol for Multihop Wireless Ad Hoc Networks. In C. Perkins, editor, Ad Hoc Networking, chapter 5, pages 139–172. Addison-Wesley, 2001.

[23] R. Karrer, A. Sabharwal, and E. Knightly. Enabling Large-scale Wireless Broadband: The Case for TAPs. In ACM HotNets 2003.

[24] J. Li, Z. J. Haas, M. Sheng, and Y. Chen. Performance Evaluation of Modified IEEE 802.11 MAC for Multi-Channel Multi-Hop Ad Hoc Network. In IEEE International Conference on Advanced Information Networking and Applications (AINA) 2003.

[25] Y. Li, H. Wu, D. Perkins, N.-F. Tzeng, and

M. Bayoumi. MAC-SCC: Medium Access Control with a Separate Control Channel for Multihop Wireless Networks. In IEEE International Conference on Distributed Computing Systems (ICDCS) Workshop 2003.

[26] A. Nasipuri and S. R. Das. Multichannel CSMA with Signal Power-Based Channel Selection for Multihop Wireless Networks. In IEEE Vehicular Technology Conference (VTC) 2000.

[27] J. Padhye, R. Draves, and B. Zill. Routing in Multi-radio, Multi-hop Wireless Mesh Networks. In ACM MobiCom 2004.

[28] C. Perkins, E. Belding-Royer, and S. Das. Ad hoc On-Demand Distance Vector (AODV) Routing. In IETF RFC 3561, July 2003.

[29] R. Rozovsky and P. Kumar. SEEDEX: A MAC Protocol for Ad Hoc Networks. In ACM MobiHoc 2001.

[30] E. Shih, P. Bahl, and M. Sinclair. Wake on Wireless: An event driven power saving strategy for battery operated devices. In ACM MobiCom 2002.

[31] J. So and N. H. Vaidya. Multi-Channel MAC for Ad Hoc Networks: Handling Multi-Channel Hidden Terminals Using a Single Transceiver. In ACM MobiHoc 2004.

[32] A. Tyamaloukas and J. J. Garcia-Luna-Aceves. Channel-Hopping Multiple Access. In IEEE International Communications Conference (ICC) 2000.

[33] S.-L. Wu, C.-Y. Lin, Y.-C. Tseng, and J.-P. Sheu. A New Multi-Channel MAC Protocol with On-Demand Channel Assignment for Mobile Ad Hoc Networks. In International Symposium on Parallel Architectures, Algorithms and Networks (I-SPAN) 2000.

[34] S. Xu and T. Saadawi. Does the IEEE 802.11 MAC Protocol Work Well in Multihop Wireless Ad Hoc Networks? IEEE Communications Magazine, pages 130–137, June 2001.

[35] Z. Tang and J. J. Garcia-Luna-Aceves.

Hop-Reservation Multiple Access (HRMA) for Ad-Hoc Networks. In IEEE Infocom 1999.

CURRICULUM VITAE

Bruce J. Perlman, Ph.D.

bperlman@unm.edu

HOME: 8409Vintage Dr. N.E. OFFICE: School of Public Administration

Albuquerque, New Mexico, 87122 Social Sciences Bldg, Room 3004

(505) 856-5026 MSC05 3100

(505) 321-3642 (cellular) 1University of New Mexico

Albuquerque, New Mexico 87131-0001

(505) 277-3313

CURRENT POSITION:

Professor, School of Public Administration, University of New Mexico, Albuquerque, New Mexico

EDUCATION:

Ph.D. in Government, Claremont Graduate School, Claremont, California,1980

B.A. in Social Science/History, California State Polytechnic University, 1972

FELLOWSHIPS, HONORS, and AWARDS:

Hatch Chair in Law and Public Administration, 2001 to 2004, University of New Mexico, School of Public

Administration, Albuquerque, New Mexico

Presidential Lecturer in Public Administration, 1988 to 1990, University of New Mexico, School of Public Administration, Albuquerque, New Mexico

Fulbright Senior Lecturer, 1986 to 1987, Law Faculty, Public Administration Career, National Autonomous University of Honduras, Tegucigalpa, Honduras

Dissertation Fellow, Haynes Foundation, 1978 to 1980, Claremont Graduate School, Claremont, California

Interdisciplinary Crime and Delinquency Fellow, Law Enforcement Assistance Administration, 1973 to 1976, Claremont Graduate School, Claremont, California

GRANTS:

Research Grant, “The “North-South Project: the Politics of Infrastructure in Eurasia and the Middle East”,

July 2005 to June 2006, International Research and Exchange (IREX), U.S. Department of State

Research Grant, “Government Reform in Central Asia”, July, 2004, International Research and Exchange (IREX), U.S. Department of State

Research Grant, “Collaboration on Government Reform and Curriculum Development with Urals Academy of Government Service”, September 2003 to December 2003, Network of Institutes and Schools of Public Administration in Central and Eastern Europe (with Gleason and Sazanova)

Research Grant, “Why So Few Women in Information Technology: A Comparative Study”, October 2001

to September 2002, National Science Foundation, Information Technology Workforce Program, Co-Principal Investigator (with R. Varma and D. Kapur)

Academic Specialist Grant in Public Administration: Municipal Development, 1994, Colombia Externado University, United States Information Agency, Bogotá, Colombia

Academic Specialist Grant in Political Science: Public Administration, 1989, Law Faculty, Public Administration Career, National Autonomous University of Honduras, United States Information Agency, Tegucigalpa, Honduras

. Perlman Curriculum Vita Page 2 of 16

Inter-American Field Research Grant, 1986, National Personnel Directorate of Ecuador, Mellon Foundation, Quito, Ecuador

REPORTED RESEARCH:

Books and Monographs

The New Institutionalism in Mexico: Essays on Mexican Public Administration, B. Perlman and J.

Pineda, (eds.), Mexico City, Mexico: University of Aguascalientes, 2004

“Large Scale Job Dislocation and Reduced Economic Growth: Market Failure or Government Failure?” (with S. Nystrom), U.S. Congress, Joint Economic Committee, September 24, 1996

Journal and Other Publications

“Introduction: Energizing Facilities,” State and Local Government Review, 44:2 May 2012, pp. 37- 39

“Social Media Sites at the State and Local Level: Operational Success and Governance Failure,” State and Local Government Review, 44:1 April, 2012, pp. 67-75

“The Great Recession and Compensation Practices: New Rules for Compensation?" State and Local Government Review, 43:3 December, 2011, pp. 221-223

“Capital Ideas: National Capitals as a Fourth Jurisdiction,” State and Local Government Review, 43:2 August, 2011, pp. 140-143

“Rethinking a Megalopolis: A Metropolitan Government Proposal for the Mexico City Metro Area,” State and Local Government Review, 43:2 August, 2011, pp. 144-150 (with J. Pineda)

“Introduction: Risks and Rewards in State and Local Collaboration,” State and Local Government Review, 43:1 2011, pp 46-48

“Governance Challenges and Options for State and Local Government,” State and Local Government Review, 42:3 2010, pp 246-257

“Creative Regionalism: Governance for Stressful Times,” State and Local Government Review, 42:2, 2010 (with J. Jimenez), pp 151-156

“Introduction: New Rules and Approaches for Succession Planning,” State and Local Government Review, 42:1, 2010, pp 48-50

“Fiscal Distress and Governance Challenges: The Perfect Storm of the Fiscal Crisis,” State and Local Government Review, 41:3, 2009, pp. 201-207

“The ARRA of Our Ways,” State and Local Government Review, 41:2, 2009, pp. 120-122 “Governance and Sustainability,” State and Local Government Review, 41:1, 2009, pp. 37-38

“Cultural Determinism versus Administrative Logic: Asian Values and Administrative Reform in Kazakhstan and Uzbekistan,” International Journal of Public Administration (with G. Gleason), 30:12, 2007, pp. 1-16 (1327-1342)

“Planeación e Integración Estratégica: Un Modelo con Estudio de Caso del Sector Judicial”, Estado, Gobierno, Gestión: Revista Chilena de Administración Pública, Diciembre, 2006 (with M. Rivera), pp. 112-124

. Perlman Curriculum Vita Page 3 of 16

“Barely Managing: Attitudes of Information Technology Professionals on Management Technique”, The Social Science Journal, (with R. Varma), 42:4, 2005, pp. 583-594

“El Estado Actual de la Ciencia Política ”, Enlace: La Revista del Colegio Nacional de Ciencias Políticas y Administración Pública de México, June, 2005

"Issues in Comparative and International Administration," Handbook of Public Administration, Jack Rabin, (ed), 3rd Edition, New York: Marcel Dekker, Public Administration and Public Policy Series (with F. Heady and M. Rivera), 2005, pp. 605-626

“Presentación,” Conflicto Político y Negociación, A. Acosta Leon and J. Pineda (eds), Mexico City, Mexico: National College of Political Science and Public Administration, 2005

“Una Revisión Institucional Sobre Cuatro Experiencias de Servicio Civil y/o Profesional de Carrera”, Servicio Profesional de Carrera, (with J. Pineda), Vol. II, núm. 4, segundo semestre de 2005

“Comparative Perspectives on Third Generation Reform: Realignment and Misalignment in Central Asian Reform Programs”, International Public Management Review, (with G. Gleason), 6:1, 2005, pp. 100¬117

“Report on the International Public Management Network Research Conference in Rio de Janeiro, Brazil,” International Public Management Review, (with E. Caperchione and T. Virtanen), 6:1, 2005, pp. 169¬178

“The Rubric of Reform in Muslim Central Asia”, PA Times, (with G. Gleason), Vol. 27, No. 8, August, 2004, pp. 8-10

“Estudio Introductorio: La Gestión Pública Local en los Estados de América Latina”, La Gestión Pública en Gobiernos Locales: Experiencias Latinoamericanas, (J. Pineda and R. Grandinetti, eds.), pp. 11-15, México: Innovación Editorial Lagares, 2004

“Algunas Lecciones en Gestión Pública Comparativa”, Revista del Instituto de Administración Pública del Estado de México, No. 110, Enero-Abril, 2004, pp. 229-243

“Seguridad Jurídica y Moralidad Gubernamental”, Revista del Instituto de Administración Pública del Estado de México, No. 57, Enero-Abril, 2004, pp. 79-102

“Managing the War Against Terror (Part 2): Emergency Response to Terrorist Incidents”, Book Review Essay, Public Performance and Management Review, Vol. 26, No. 4, June, 2003, pp. 441-448

“Managing the War Against Terror: A Policy Framework for the United States”, Book Review Essay, Public Performance and Management Review, Vol. 26, No. 2, December, 2002, pp. 214-220

“Understanding and Improving Engineering Practice”, IEEE Technology and Society, (with R. Varma), 21:1, 2002, pp. 40-47

"Un Nuevo Modelo para la Formulación de Estrategias en la Planificación", Carta Económica Regional, Año 14, No. 78, (with M. Rivera), Octubre-Diciembre 2001, pp. 25-34

“What’s Wrong and what should be done: Comments on the Case Study, Who’s on First,” Public Performance and Management Review, Vol. 25, No. 1, September, 2001

. Perlman Curriculum Vita Page 4 of 16

"Bridging the Cultural Chasm: Improving Collaboration and Cooperation between the Computer and Social Sciences", Proceedings of the International Symposium on Technology and Society (with R. Varma), IEEE Catalog No. PR01209, 2001, pp.19–27

"Teaching Engineering Ethics", Proceedings of the American Society of Engineering Education Annual Conference & Exposition, Session 2793 (with R. Varma), June, 2001

"Institutional Constraints on Market Liberalization Policies in Cuba", Policy and Management Review, (with M. Rivera), Volume 1, No. 1, 2000, pp. 64-91

“Development and Reform of Post Sandinista Nicaraguan Public Administration”, Research in Public Administration, Volume 5, J. Perry (ed), JAI Press, Inc., 1999, pp. 189-231

“The Emergence of Collective Bargaining in the Guatemalan Public Sector: A Case of ‘Union Bulwarism’?” Public Personnel Management, (with T.Z. Reeves), Vol. 27, No.2, Summer, 1998

“Nicaragua: From Patrimonialism to Positivism: Assessing Past and Planned Reforms and a Theory of Configuration”, Proceedings, Conference on Civil Service Systems in Comparative Perspective, School of Public and Environmental Affairs, University of Indiana, Bloomington (April 5-8, 1997)

"La Evolución de las Gestiones Realizadas por Estados Unidos Para Controlar el Narcotráfico a Nivel Internacional Y Sus Repercusiones en Latinoamérica: Una Investigación Preliminar", (with G. LaFree), Chapter 4, La Criminología en America Latina, Birkbeck & Martínez, (eds.), Merida, Venezuela: Universidad de los Andes, Consejo de Desarrollo Científico, Humanístico y Tecnológico, 1992

"Teaching Methods in Public Administration Revisited,"(with T. De Young), Policy Studies Review, Vol. 8, No. 4, summer 1989, pp 852-858

"Bandidos: The Varieties of Latin American Banditry," Book Review, Criminal Justice Review, Vol. 13, No. 1, spring, 1988, pp. 73-75

"Modernizing the Public Service in Latin America: Paradoxes of Latin American Public Administration," International Journal of Public Administration, Vol. 12 No. 4, 1988, pp.671-704

"Who Was That Masked Man: Local Law Enforcement in New Mexico and the Southwest," Urban Resources, Vol. 4, No. 3, spring, 1987

"Ethical Dimensions of Public Policy: A Conceptual Framework," (with D. Rosenthal), Polity Vol. XIX, No. 1, autumn, 1986, pp. 56-73

"Tools for Program Evaluation", Proceedings, 9th National Conference on Teaching Public Administration, (with T. De Young), May 1, 1986, New York, New York

"Microcomputers in Public Administration: Prerequisite to Productivity", Public Productivity Review (with T. De Young), Vol. IX, Number 2-3, Summer/Fall, 1985, pp. 247-259

"Teaching Research Methods in Public Administration" Teaching Political Science: Politics in Perspective (with T. De Young), Vol. 11, No. 2, winter 1983-1984: 63

"A Political Economy Approach to Criminal Justice Education: Concerns for Study and Teaching," Proceedings, 7th National Conference on Teaching Public Administration, March 8, 1984, Albuquerque, New Mexico

. Perlman Curriculum Vita Page 5 of 16

"Teaching Methods for Teaching Methods in Public Administration," Proceedings, 6th National Conference on Teaching Public Administration (with T. De Young), March 1, 1983, Miami, Fl.

Technical Reports

“Actualización de la estructura orgánica del MEC: Avance en el fortalecimiento organizacional del MEC,” Programa para el desarrollo la educación de Guinea Ecuatorial (PRODEGE), Ministerio de Educación y Ciencias de Guinea Ecuatorial, April 2011

“Annual Evaluation Report,” WESST Corp JOLI Grant, U.S. Department of Health and Human Services, October, 2004

“Monitoring Report,” WESST Corp JOLI Grant, U.S. Department of Health and Human Services, April, 2003

“Hardmark Telecommunications Grant: Project Review and Evaluation,” New Mexico Department of Labor, Albuquerque New Mexico, March, 2003

“Statewide Vulnerability Assessment,” State of New Mexico Department of Public Safety Office of Emergency Services and Security, January, 2003

“Brief Comparison: Functions of Elected At Large County Officials and Appointed County

Administration,” Urban County Charter Commission, County of Bernalillo, April, 2001 to May, 2001

“Audit Response for NMDOL United States Department of Labor,” Office of Inspector General, New Mexico Department of Labor, July, 2001 to August, 2001

“Large Scale Job Dislocation and Reduced Economic Growth: Market Failure or Government Failure?” U.S. Congress, Joint Economic Committee, September 24, 1996 (with S. Nystrom)

“Guatemala: Modernization of the Public Sector and Human Resource Management,” International Bank for Reconstruction and Development (World Bank), Washington, D.C., January, 1996

“Re-engineering the Costa Rican Central Government: Action Brief,” Sigma One Corp. and USAID, Costa Rica, San Jose, Costa Rica, March, 1995

“Technical Assistance Action Plan, Nicaragua Civil Service Reform Technical Assistance Credit,” International Bank for Reconstruction and Development (World Bank), Washington, D.C., October, 1994

“Nicaraguan Civil Service Review, Modernization of the State, Civil Service Component,” International Bank for Reconstruction and Development (World Bank), Washington, D.C., August, 1994

"Evaluation of Southeast Regional Biomass Program, Chapter IV: Lessons Learned," Sandia National Laboratories and DOE/EE/OPA, Albuquerque, New Mexico, May, 1994

"Analysis of EE Sector Program Evaluation Indicators, Key Sector, and Key Written Sources," Sandia National Laboratories and DOE/EE/OPA, Albuquerque, New Mexico, March, 1994

"Estudio de Instituciones Claves," Development Training Project, Creative Associates International Inc. and USAID/GDO, Nicaragua, October, 1994

"Manual de Funciones del Ministerio de Educación," Education and Human Resources Training Support, Juarez and Associates and USAID/GDO, Nicaragua, March, 1993

. Perlman Curriculum Vita Page 6 of 16

"The Future of Municipal Service Delivery in Phoenix and Maricopa County", Arizona State University, School of Public Affairs, April, 1992

"Sub Sector Assessment for Basic Education", Academy for Educational Development and USAID/GDO, Nicaragua, August, 1991

"A Paper on Issues in the Study of Problems Concerning and the Design of Programs for the Reintegration of Salvadoran Ex Combatants into Civilian Life", Creative Associates International Inc. and USAID/IRD, El Salvador, December, 1990

"A Report and Recommendations Including Training Plan on Guatemala Public Administration, Financial Management and Anti-corruption," USAID/HRD Guatemala, June, 1990

"Citizen Satisfaction with Police Service Delivery," University of New Mexico, Institute for Criminal Justice Studies, May, 1990

"Drug Enforcement on the Argentine Border: Report and Recommendations," Procuración de la Nación y Policía Federal de Argentina, August, 1989

"Training Plan for APRE Leadership," USAID/HRD, El Salvador, July, 1987

"Estudio Administrativo del Poder Judicial de Honduras, USAID/ODI, Honduras," December, 1986

"Program Summary: Programa de Capacitación en Administración Pública, Grupo Ecuatoriano", USAID/HRD, Ecuador" August, 1985

"Management Information Systems Evaluation Study," New Mexico State Highway Department, May, 1984.

"Honor Farm Feasibility Study," Bernalillo County Detention Center, December, 1983

SELECTED PAPERS AND PRESENTATIONS:

Academic Papers and Presentations

Paper, “La Agenda Pendiente en el Sistema Institucional del Estado Mexicano ”, Panel Convener, “Análisis Institucional, Profesionalización, y Modernización del Sector Público en México”, X International Conference on Reform of the State and Public Administration, Latin American Center for Development Administration, Santiago, Chile, Pública, November 18-21, 2005

Paper, “Comparative Perspectives on Third Generation Reform: Realignment and Misalignment in Central Asian Reform Programs”, Conference of the International Public Management Network at the Getulio Vargas School of Public and Business Administration, Rio De Janeiro, Brazil, November 17-19, 2004

Paper, “Formalismo y Pragmatismo en las Administraciones Públicas en el Continente Americano: Filosofía y Práctica en el Norte y en el Sur”, IX International Conference on Reform of the State and Public Administration, Latin American Center for Development Administration, Instituto Nacional de Administración Pública, Madrid, Spain, November 2-5, 2004

Paper, Una revisión inicial a la nueva Ley del Servicio Profesional de Carrera en la Administración Pública Federal de México: Oportunidades y Desafíos en la Medición de Resultados, VIII Internacional Conference on Reform of the State and Public Administration, Latin American Center for Development Administration, Panama City, Panama, November 28-31, 2003

. Perlman Curriculum Vita Page 7 of 16

Paper, “Algunas Lecciones en Gestión Pública Comparativa”, VIII Internacional Conference on Reform of

the State and Public Administration, Latin American Center for Development Administration, Panama City, Panama, November 28-31, 2003

Paper, Hazards vs. Targets in the State of New Mexico: Fitting Local Problems to National Priorities in the Homeland Security Planning Process, 64th Annual Conference of the American Society of Public Administration, Washington, D.C. March 14-19, 2003

Paper, “Barely Managing: Perceptions of Computer Scientists on Management and the Social Sciences in Public Institutions”, 44th Annual Conference of the Western Social Science Association, Albuquerque, New Mexico, April 10-13, 2002

Paper, “Managing Computer Scientists in Government Service”, 63rd Annual Conference of the American Society of Public Administration, Phoenix, Arizona, March 23-26, 2002

Paper, Seguridad Jurídica y Capital Social, 6th Ibero-American Juridical Conference, National Judicature Council, San Salvador, El Salvador, September 3-6, 2000

Paper, "Managing on the Border", Panel Convener, "Managing on the Border," 53rd National Conference, American Society for Public Administration, Chicago, Illinois, April, 1992

Panel Convener, "Comparative Public Administration" Western Social Science Association, 33rd Annual Conference, Denver, Colorado, April, 1992

Paper, "Issues in Multi-cultural Training," Western Social Science Association, 31st Annual Conference, Portland, Oregon, April, 1990

Paper, "Police Professionalism: Guild Ethics or Professional Ethics?" 50th National Conference, American Society for Public Administration, Miami, Florida, April, 1989

Paper, "Technical Assistance in Civil Conflict Areas," 50th National Conference, American Society for Public Administration, Miami, Florida, April, 1989

Presentation, "What to Do With Excellence after You Find It," New Mexico Municipal League, City Managers' Section, Taos, New Mexico, April, 1989

Presentation, "Performance and Performing," International Municipal Clerks National Conference, Albuquerque, New Mexico, January, 1989

Paper "Justice under the Gun: Judicial Administration in El Salvador, "Panel Convener, "Administration in

Conflict," 49th National Conference, American Society for Public Administration, Portland, Oregon, April, 1988

Paper, "Law Enforcement Firearms Training, "Panel Convener, "Democracy Under the Gun: A Round Table on Firearms Policy, Administration and Training." 48th National Conference, American Society for Public Administration, Boston, Massachusetts, April, 1987

Paper and Computer Demonstration, "Using Microcomputers to Make Decisions," 9th Annual National Conference on Teaching Public Administration, White Plains, New York, May, 1986

Paper, "The Role of Local Police in Combating Terrorism", 47th National Conference, American Society for Public Administration, Anaheim, California, April, 1986

. Perlman Curriculum Vita Page 8 of 16

Paper, "Varieties of Violence", Panel Convener, "Varieties of Violence", 47th National Conference, American Society for Public Administration, Anaheim, California, March, 1986

Discussant, "Air Quality in Albuquerque", New Mexico Chapter, American Society for Public Administration, Albuquerque, New Mexico, February, 1986

Paper, "La Evolución de las Gestiones Realizadas por Estados Unidos Para Controlar el Narcotráfico a Nivel Internacional Y Sus Repercusiones en Latinoamérica: Una Investigación Preliminar," Seminario Internacional, La Criminología en America Latina: Balance y Perspectivas, Universidad de los Andes, Mérida, Venezuela, November, 1985 (with G. LaFree)

Paper, "New Techniques in Public Administration," Panel Convener, "New Techniques in Public Administration", 8th Annual Conference on Teaching Public Administration, St. Louis, Missouri, April, 1985

Paper, "Computers in Law Enforcement: A Review," Western Political Science Association, Annual Conference, Las Vegas, Nevada, March, 1985

Paper, "Law Enforcement Intelligence Units: The Myth of Strategic Intelligence, Western Political Science Association, Annual Conference, Las Vegas, Nevada, March, 1985

Paper, "Police Assessment in One Southwestern City," American Society for Public Administration, National Conference, Denver, Colorado, April, 1984

Paper, "Reasons for Joining the Police Reserves: Experience in One Southwestern State," Western Social Science Association, 26th Annual Conference, San Diego, California, April, 1984

Paper, "Education and Performance in Assessment Centers: Experience in One Southwestern Police Department," Western Social Science Association, 26th Annual Conference, San Diego, California, April, 1984

Paper, "Ethical Dimensions of Public Policy and Public Administration," Southwestern Political Science Association Annual Conference, Fort Worth, Texas, March, 1984 (with D. Rosenthal)

Paper, "Public Impact on Corrections," New Mexico Association of Criminal Justice Professionals Annual Conference, Albuquerque, New Mexico, October, 1983

Paper, "Care and Feeding of the Violent Juvenile Offender," American Society for Public Administration, Region VIII Conference, Phoenix, Arizona, April, 1983

Paper, "Collective Bargaining in New Mexico Corrections: The Strange Case of AFSCME vs. NMCWA," Western Social Science Association, 25th Annual Conference, Albuquerque, New Mexico, April, 1983

Paper, "The Moral Fiction of Expertise: The Cases of the Policy and Management Sciences," Western Social Science Association, 25th Annual Conference, Albuquerque, New Mexico, April, 1983 (with D. Rosenthal)

Paper, "Reaganomics and Corrections: Correctional Policy Making In An Era of New Federalism" American Society for Public Administration, Region VIII Conference, Denver, Colorado, October, 1982. Paper, "Productivity Enhancing Programs: Salary Structure for Rewarding Exceptional Performance," International Conference, San Diego, California October, 1982

. Perlman Curriculum Vita Page 9 of 16

Paper, "The Future of Corrections," Panel Convener, "The Future of Corrections", American Society for Public Administration Region X and Western Governmental Research Association Joint Annual Conference, Anaheim, California, October, 1982

Invited Addresses

Invited Address, “Políticas Públicas y Educación Internacional ,” IV Annual Latin American Network

Conference, Puebla, México, September 2, 2011

Invited Address, “Transformación organizacional en el estado contemporáneo,” El congreso

internacional en gerencia de proyectos y mejoramiento organizacional, EAN University, Bogotá, Colombia, September 24, 2010

Invited Address, El Estado y Las Políticas Públicas, Bases Para un Proyecto de Nación, Conferencia

Inaugural de la Cuarta Promoción, Maestría en Políticas Públicas, Universidad Rafael Landivar, Guatemala City, Guatemala, July 7, 2003

Invited Address, El Estado Actual de la Ciencia Política, Semana de Administración Pública, Escuela Nacional de Estudios Profesionales UNAM-Acatlán, México, June 5, 2003

Invited Address, Conferencia Magistral, Gerencia Pública y Rediseño Institucional en las Administraciones Públicas”, Segundo Congreso Internacional en Ciencia Política y Administración Pública, Colegio Nacional de Ciencia Política y Administración Pública de México, Mexico City, Mexico, November, 13-14, 2002

Invited Address, Algunas Lecciones sobre Administración Pública Comparada, Seminar on Comparative Methodology in Public Management, Universidad de Chile, Santiago, Chile, August26-28, 2002

Invited Address, La Dicotomía entre la Política y la Administración, Universidad de Guadalajara, Guadalajara, México, July, 1994

Invited Address, "The Philosophy of Social Science Methods" USIA Guatemala, 1st USAC-UNM Social Science Seminar, Guatemala, February, 1989

Invited Address, "U.S. Drug Enforcement Policy, 3rd U.N. Conference on Trans-national Crime, Albuquerque, New Mexico, November, 1988

Invited Address, "La Lucha Contra Las Drogas," Procuración de la Nación and Policía Federal de la Republica, Seminario Sobre Narcotráfico, Buenos Aires, Argentina, May, 1988

Invited Address, "Power, Politics, and Professional Groups, New Mexico Dietician Association Annual Meeting, Albuquerque, New Mexico, April, 1986

Invited Address, "Market Research", Work Unlimited Marketing Workshop, Albuquerque, New Mexico, February, 1986

Invited Address, "How to Fail Miserably with Your Staff", New Mexico Association of Counties, Treasurers Affiliate Legislative Conference, Santa Fe, New Mexico, January, 1986

Invited Address, "Burnout: Don't Make an Ash of Yourself", New Mexico Municipal League Annual Conference, Las Cruces, New Mexico, September, 1985

ACADEMIC AND PROFESSIONAL EXPERIENCE:

Chief Financial Officer, Office of the State Treasurer, State of New Mexico, Santa Fe, New Mexico, July,

2009 to January, 2010 (TDY from Governor’s Office; on leave from UNM)

. Perlman Curriculum Vita Page 10 of 16

Deputy Chief of Staff, Office of the Governor, State of New Mexico, Santa Fe, New Mexico, July, 2008 to July, 2009 (TDY to State Treasurer in July 2009; on leave from UNM)

Chief Administrative Officer, City of Albuquerque, Albuquerque, New Mexico, December, 2005 to June 2008 (on leave from UNM)

Director, School of Public Administration, University of New Mexico, May, 2004 to December, 2005

Chief of Party and Resident Advisor, Project for Reform of the Justice Sector, Inter American Development Bank and Associates for Rural Development, May, 1999 to July, 2000 (on leave from UNM)

Interim Director, School of Public Administration, University of New Mexico, May, 1994 to January, 1996

Visiting Associate Professor, School of Public Affairs, Arizona State University, January, 1991 to June, 1992

Associate Professor, School of Public Administration, University of New Mexico, August, 1989 to the January, 2006 (took leave to fill other posts)

Associate Director for Programs in Public Administration, Office of International Technical Cooperation, University of New Mexico, July, 1988 to July, 1990

Associate Director, Division of Public Administration, University of New Mexico, August, 1988 to August, 1989

Director, Master of Arts in Public Administration in Spanish, Division of Public Administration, University of New Mexico, January, 1987 to August, 1988

Assistant Professor, Division of Public Administration, University of New Mexico, August, 1982 to May, 1988

Academic Coordinator, Programa de Capacitación en Efectividad Organizacional, Office of International Technical Assistance, University of New Mexico, May, 1986 to February, 1987

Project Manager, A.C.O.P.S. Integration Project, Albuquerque Police Department, January, 1984 to July, 1985

Academic Coordinator, Programa de Capacitación en Administración Pública, Latin American Programs

in Education and Division of Public Administration, University of New Mexico, May, 1985 to August, 1985

Director, Policy and Systems Development, Albuquerque Police Department, January, 1983 to January, 1984 (on leave from UNM)

Lecturer, California State University, Dominguez Hills, September, 1981 to June, 1982

Program Manager, University of Southern California, June, 1980 to August, 1981

Adjunct Faculty, California State University, Dominguez Hills, September, 1979 to June, 1981

Research Associate, Executive Management Program: Humanities Component, Claremont Graduate School. September, 1977 to June, 1979

. Perlman Curriculum Vita Page 11 of 16

Evaluation Consultant, Noble and Associates, August, 1976 to September, 1977

Teaching Assistant, Sociology Department, Pitzer College, Claremont Colleges, September, 1975 to June, 1976

Director, Project Upstream, Los Angeles Unified and Pomona Unified School Districts, September, 1974 to June, 1975

CONSULTING AND CONTRACT RESEARCH:

International:

Academy for Educational Development, PRODEGE, Organizational Strengthening Consultant for the

Ministry of Education of Equatorial Guinea, October, 2010 to September, 2011

Inter American Development Bank and Associates for Rural Development, Project for Reform of the

Justice Sector, Chief of Party and Resident Advisor, May, 1999 to July, 2000 (on leave from UNM)

The International Bank for Reconstruction and Development (World Bank) and the Government of Guatemala, Ministry of Finance, Department of External Financing, Guatemala: Modernization of the Public Sector and Human Resource Management, Public Administration Specialist. Guatemala, Guatemala, July, 1995 to December, 1995

USAID Costa Rica, Office of General Development and Sigma One Corporation, Public Administration Specialist for State Reform for The Technical Secretariat of State Transformation and the Ministry of Planning, June, 1995 to September, 1995

The International Bank for Reconstruction and Development (World Bank), Nicaragua Reform of the State Project, Civil Service Modernization Component, Public Administration Specialist, Managua, Nicaragua, May, 1994 to August, 1994

USAID Nicaragua, Office of General Development, and Juárez and Assoc., Inc., Public Administration Specialist, BASE Project, Managua, Nicaragua, January, 1994 to April, 1994

USAID Nicaragua, Office of General Development, and Creative Assoc. Int'l., Inc., Public Administration Specialist, Development Training Project, August, 1993 to October, 1993

USAID Guatemala, Office of Education and Human Resources and Juárez and Assoc., Inc., Chief of Party and Education Administration Specialist, Rural Primary Education Improvement Project Final Evaluation, Guatemala City, Guatemala, January, 1993 to April, 1993

USAID Nicaragua, General Development Office, and Juárez and Associates Inc., Administration

Specialist, Functions Manual for Ministry of Education. Managua, Nicaragua, September, 1992 to March, 1993

USAID Costa Rica, Projects Development Office, and Sigma One Corporation, Public Law and Administration Consultant, Costa Rica State Public Sector Reform Project. San Jose, Costa Rica, November, 1992

USAID Guatemala, Office of Education and Human Resources, and Creative Associates International, Inc., Education Administration Specialist, Mid-term Evaluation of the Basic Education Strengthening Project, Guatemala City, Guatemala, July, 1992 to September, 1992

USAID Nicaragua, General Development Office, and Juárez and Associates Inc., Institutional Strengthening of the Nicaraguan Ministry of Education, Managua, Nicaragua, June, 1991 to March, 1992

. Perlman Curriculum Vita Page 12 of 16

USAID Nicaragua, General Development Office, and Juárez and Associates Inc., Project Paper Team Leader/Public Administration Specialist, Project Paper. Managua, Nicaragua, December, 1991 to January, 1992

The International Bank for Reconstruction and Development (World Bank), El Salvador Social Sector Project, Specialist in Public Administration and Organizational Development. San Salvador, El Salvador, October, 1991

USAID Nicaragua, General Development Office, and Juárez and Associates Inc., Sub Sector Assessment in Basic Education. Managua, Nicaragua, February, 1991 to May, 1991

USAID El Salvador, Office of Infrastructure and Rural Development, and Creative Associates International Inc., Study on the Reintegration of Ex Combatants into Salvadoran Civil Society, San Salvador, El Salvador, November, 1990 to December, 1990

USAID Guatemala, Office of Human Resource Development, Development Training and Support

Program, Materials Inventory and Operations Review of Controllers General Offices of Costa Rica and Spain. San Jose, Costa Rica; Guatemala, Guatemala, June, 1990 to July, 1990

USAID Guatemala, Office of Human Resource Development, Development Training and Support Program, Diagnosis and Needs Assessment for Administrative Reform for the Controller General's Office of Guatemala, Guatemala, Guatemala, November, 1989 to June, 1990

USAID El Salvador, Infrastructure and Rural Development Program, Administrative Reform and

Management Training and Technical Assistance Program for Engineers, Albuquerque, New Mexico; Antigua, Guatemala; San Salvador, El Salvador, January, 1989 to June, 1989

USIA Costa Rica and USAID Regional Administration of Justice Office, Court Administration Course, Albuquerque, New Mexico, May, 1989 to August, 1989

USIA Guatemala, University of San Carlos, Guatemala, UNM, Partnership Program in Political Science, Guatemala, Guatemala; Albuquerque, New Mexico, January, 1989 to December, 1989

Procuración de la Nación y Policía Federal de Argentina, Study on Organizational Cooperation in Drug

Enforcement on the Argentine-Bolivian Border. Buenos Aires, Argentina, Salta, Argentina; Positos, Bolivia, August, 1989

USAID Guatemala, Office of Human Resources, National Institute of Public Administration of Guatemala, Training Program for Executives. Guatemala, Guatemala, July, 1988

USAID El Salvador, Office of Education and Training, and UNM Office of International Technical Assistance, El Mejoramiento de la Productividad en el Sector Público, una Evaluación y Taller de Trabajo. San Salvador, El Salvador, June, 1987 to July, 1987

USAID El Salvador, Office of Democratic Initiatives, and Proyecto Ilanud-FIU-El Salvador, Estudio Sobre El Órgano Judicial en El Salvador. San Salvador, El Salvador, May, 1987 to June, 1987

USAID El Salvador, Office of Education and Training and UNM Office of International Technical Assistance, Proyecto para Revitalización de Educación. San Salvador, El Salvador, March, 1987 to April 1987

USAID El Salvador, Office of Education and Training, and, UNM Office of International Technical

Assistance, Program in Organizational Effectiveness, San Salvador, El Salvador; Albuquerque, New Mexico, May, 1986 to February, 1987

. Perlman Curriculum Vita Page 13 of 16

USAID Honduras, Office of Democratic Initiatives, and Proyecto Ilanud-FIU-Honduras, Estudio de La Administración de Justicia, Estudio Administrativo. Tegucigalpa, Honduras, November, 1986 to December, 1986

USAID Ecuador, Office of Education and Training, and UNM Latin American Programs in Education and Division of Public Administration, Program in Public Administration, Quito, Ecuador; Albuquerque, New Mexico, May, 1985 to August, 1985

Domestic:

City of Albuquerque, Albuquerque Fire Department, Update and Review of Standard Operating Guidelines,

October, 2004 to December, 2005

City of Albuquerque, Environmental Health Department, Metropolitan Medical Response System Consultant, May, 2004 to December, 2005

Techonex, Program Evaluator for the Hardmark Telecommunications Implementation Grant, United States Department of Labor, October, 2003 to December, 2005

City of Albuquerque, Office of Emergency Preparedness, Emergency Planning and Management Consultant, July, 2002 to January, 2005

State of New Mexico Department of Public Safety Office of Emergency Services and Security, Consultant for Statewide Vulnerability Assessment, July, 2002 to January, 2003

Women’s Economic Self Sufficiency Team Corp. (WESST Corp), Program Evaluator for the JOLI Micro

enterprise Development Grant, United States Department of Health and Human Services, Albuquerque New Mexico, December, 2002 to present

Central Area Workforce Investment Board, Workforce Investment Act, Grants Administration Consultant, Albuquerque New Mexico, October, 2002 to December 2003

New Mexico Department of Labor, Program Evaluator for the Hardmark Telecommunications Demonstration Grant, United States Department of Labor Albuquerque New Mexico, September, 2002 to March, 2003

State of New Mexico Department of Public Safety Office of Emergency Services and Security, Consultant for Statewide Vulnerability Assessment, July, 2002 to January, 2003

Urban County Charter Commission, County of Bernalillo, Consultant for Functional Comparison of Elected and Appointed Officials, April-May, 2001

New Mexico Department of Labor, Consultant for Preparation of Audit Response to United States Department of Labor, Office of Inspector General, July-Aug, 2001

Sandia National Laboratories and the Federal Aviation Administration, Flight Services Division. Consultant for Organizational Systems, Aviation Technical Oversight System, September, 1997 to February, 1999

Sandia National Laboratories and the Federal Highway Administration, Office of Motor Carriers, Consultant for Field Research and International Administration, Advanced Technologies for International, Intermodal, Ports of Entry, July, 1996 to February 1997

. Perlman Curriculum Vita Page 14 of 16

Sandia National Laboratories and Department of Energy, Conservation and Renewable Energy, Office of Planning and Evaluation, Consultant in Evaluation, July, 1993 to June, 1994

New Mexico Municipal League and New Mexico Association of Counties, Mediator and Chair, Joint Executive Board Meeting for Legislative Agenda Setting, December, 1984 to 1990

City of Albuquerque, New Mexico, Albuquerque Police Department, Citizen Satisfaction with Service Delivery Study. January to 1990, May, 1990

Council for International Development, Washington, D.C., Proposal for Afghanistan Opium Reduction Project, April, 1990

New Mexico State Legislative Finance Committee, Planning Retreat, February, 1990

City of Santa Fe, New Mexico, City Council Planning Retreat, February, 1990

City of Silver City, New Mexico, City Manager and Department Directors' Planning Retreat, July, 1989

New Mexico Municipal League, Statewide Training Needs Assessment, February, 1989 to August, 1989

Silva & Silva, El Paso, Texas, Expert Witness, Pérez v. F.B.I., December, 1989 to May, 1989

City of Albuquerque, Office of Senior Affairs, Strategic Planning Review, May, 1986 to August, 1986

City of Albuquerque, New Mexico, Environmental Health Division, Project Assistance for Inspections and Maintenance Program. July, 1985

New Mexico Attorney General's Office, Expert Testimony, Childers v. Hobbs. April, 1985, to June, 1985

County of Bernalillo, New Mexico, Bernalillo County Detention Center, Honor Farm Feasibility Study. May, 1985

Communication Companies International, Malibu, California, Corporate Reorganization Study and Executive Development Program, June, 1984

Ergodotics Inc., La Verne, California, Executive Development Program, June, 1983

New Mexico State Highway Department, Management Information Systems Evaluation Study, October, 1983 to January, 1984

UNIVERSITY, PROFESSIONAL, AND COMMUNITY SERVICE:

Member, Impact Fee Review Committee, City of Albuquerque, Office of the Mayor, October 2010 to

present

Editor, Governance Matters Section, State and Local Government Review, January, 2008 to present Member, Faculty Senate Government Relations Committee August, 2010, to present

Member, Editorial Board, Public Performance and Management Review, May, 2003 to present Editor in Chief, Journal of Public Administration Education, August 2004 to July 2007

Member, Executive Board, Section on International and Comparative Administration, American Society for Public Administration, November, August 2004 to September, 2007

. Perlman Curriculum Vita Page 15 of 16

Associate Editor for Latin America, International Public Management Review, October, 2004 to November, 2007

Associate Editor for Latin America, Encyclopedia of Public Administration, September, 2004 to November, 2007

Member, International Awards Committee, American Society for Public Administration, June, 2004 to July 2007

Faculty Senator, School of Public Administration, University of New Mexico, August, 2003 to September, 2004

Member, Faculty Senate Government Relations Committee August, 2002, May to 2004

Member, Editorial Board, Public Performance and Management Review, May, 2003 to present

Faculty Intern, Office of the Vice President for Student Affairs, University of New Mexico, January, 2003 to December, 2003

Member, Executive Board, Section on Public Administration Research, American Society for Public Administration, May, 2002 to present

Member, National Conference Program Committee, American Society for Public Administration, June, 2002 to March, 2003

Representative, Faculty Senate Curriculum Committee, School of Public Administration, University of New Mexico, August, 2001 to May, 2003

Chair, Mayor’s Transition Team for the Albuquerque Police Department, November, 2001 to December, 2002

Member, City of Albuquerque Mayor's Law Enforcement Transition Team, December, 1996 to February, 1997

Member, Community Oriented Policing Steering Committee, City of Albuquerque, January, 1995 to December, 1997

Member, Faculty Senate Graduate Committee, September, 1993 to May, 1997

Member, President's Task Force on International Issues, American Society for Public Administration, September, 1992 to present

Member, Office of U.S. Senator Jeff Bingaman, U.S. - Mexico Border Business Task Force, September, 1990 to December, 1990

Member, City of Albuquerque Mayor's Law Enforcement Transition Team, December, 1989 to February, 1990

Member, City of Albuquerque Mayor's Task Force on the Police, January, 1988 to May 1988

Chairman of the Board and Faculty Associate, UNM Institute for Criminal Justice Studies, January, 1987 to December, 1989

. Perlman Curriculum Vita Page 16 of 16

Member, National Conference Program Committee, American Society for Public Administration, March, 1986 to March, 1988

National Conference Program Chair, American Society for Public Administration, Section on Criminal Justice Administration, March, 1986 to March, 1987

Member, Publications Subcommittee, Latin American Institute, University of New Mexico, August, 1986 to May, 1988

Member, University Budget Review Committee, August, 1985 to May, 1987

Regional Representative, Western Governmental Research Association; Judge, Samuel C. May Research Paper Award, January, 1985 to December, 1985

Steering Committee, City of Albuquerque, Mayor's Task Force on Alcohol and Drug Abuse, January, 1985 to June, 1985

Executive Council Member, New Mexico Chapter, American Society for Public Administration. January, 1984 to May, 1986

National Program Committee, American Correctional Association, January, 1984 to December, 1984 Vice President, New Mexico Correctional Association. January, 1983 to June, 1984

Member, Bernalillo County Adult Local Selection Panel New Mexico State Department of Corrections, January, 1983 to May, 1984

Executive Board Member, Albuquerque Police Department Reserve Corps, March, 1983 to June, 1985 Member, Educational Leave Committee, Albuquerque Police Department, January, 1983 to June, 1987

University Speakers Bureau Talks: for Kiwanis, Optimists, Exchange Clubs, N.M. Computer Society and others

LANGUAGES:

Speaking Fluency: English and Spanish

Abstrak . Mach Zhender Inferometer menggunakan Semiconductor Amplifier optik ( SOA ) dan coupler optik . Perangkat ini digunakan untuk menghasilkan fungsi logika ( AND, XOR ) & multiplexer dan Encoder yang diperoleh menggunakan perangkat ini di Optical Pohon Arsitektur . Simulasi Encoder dan Multiplexer dilakukan pada tingkat 10 Gbit / s dan keduanya disimulasikan untuk input kombinasi logis yang berbeda . Simulasi menunjukkan bahwa Perangkat ini cocok untuk beroperasi pada bit rate yang lebih tinggi dan juga untuk entitas logis yang berbeda .

Banyak aliran data kecepatan rendah dapat di-multiplexing menjadi satu kecepatan tinggi aliran melalui waktu Optical division multiplexing ( OTDM ) , sehingga setiap saluran masukan mentransmisikan data dalam ditugaskan slot waktu . Tugas ini dilakukan dengan cepat beralih multiplexer ( mux ) . The routing data yang berbeda aliran di akhir link TDM dilakukan oleh switch demultiplexer ( demux ) dan demultiplexer ini digunakan menggunakan MZI saklar terdiri semikonduktor amplifier optik ( SOA ) dan optik coupler . Dalam bab empat channel OTDM disimulasikan pada 40 Gbit / s dan selanjutnya diselidiki dampak dari kekuatan sinyal , lebar pulsa dan kekuatan sinyal kontrol pada BER . Indeks Situs Syarat . Semua saklar optik , Mach – Zehnder interferometer ( MZI ) , Semiconductor amplifier optik ( SOA ) , Skema Switching , analisis Spectrum .

I PENDAHULUAN

Dalam era informasi , teknologi melihat tanpa henti permintaan untuk jaringan kapasitas yang lebih tinggi dengan biaya lebih rendah . Teknologi komunikasi optik telah berkembang pesat ke mencapai kapasitas transmisi yang lebih besar dan transmisi lebih lama jarak . Untuk itu kecepatan data tersebut dapat dicapai jika data tetap dalam domain optik menghilangkan kebutuhan untuk mengkonversi sinyal optik . Oleh karena itu , untuk berhasil dapat mencapai kecepatan data yang lebih tinggi , jaringan optik canggih akan mewajibkan semua pemrosesan sinyal ultra cepat optik seperti konversi panjang gelombang , logika optik dan aritmatika pengolahan, fungsi add -drop , dll Berbagai arsitektur , algoritma , operasi logis dan aritmatika telah diusulkan di bidang optik / komputasi optoelektronik dan pemrosesan paralel dalam tiga dekade terakhir . nonlinier optik lingkaran cermin ( NOLM ) memberikan dukungan besar untuk optik switching yang berbasis semua logika optik dan pengolahan aljabar di mana mekanisme switching didasarkan pada serat Kerr nonlinier .

Solusi yang lebih efisien dan kompak dapat diwujudkan oleh semua switching optik dalam semikonduktor amplifier optik ( SOA ) di mana koefisien non linear jauh lebih tinggi . Konfigurasi switching berbasis Berbagai SOA telah menunjukkan sebelumnya seperti Tetrahertz asimetris optik demultiplexers ( kodok ) , inferometers nonlinier ultra-cepat ( Unis ) dan Mach - Zehnder inferometers ( MZIs ) . Diantara topologi yang berbeda , switch MZI monolithically terintegrasi merupakan solusi yang paling menjanjikan karena mereka kompak ukuran, stabilitas termal dan daya rendah . Dalam komputasi optik , sistem interkoneksi optik adalah primitif yang merupakan berbagai algoritma optik dan arsitektur . Arsitektur pohon optik ( OTA ) juga mengambil peran penting dalam hal ini . Jadi dalam era teknologi berubah dengan cepat kami mewakili skema alternatif baru yang mengeksploitasi keunggulan dari kedua SOA - MZI dan OTA , untuk implementasi dari semua logika dan aritmatika operasi paralel optik biner data.

1,1 Mach Zehnder Inferometer

The Mach - Zehnder Inferometer adalah alat yang digunakan untuk menentukan pergeseran fasa yang disebabkan oleh contoh kecil yang ditempatkan di jalan salah satu dari dua balok collimated dari dua cahaya koheren sumber . A Mach - Zehnder Interferometer dibuat dari dua skrup dihubungkan dengan lengan panjang optik yang tidak sama . itu Mach - Zehnder Interferometer

memiliki dua port input dan dua output port . Lampu dibagi dalam dua lengan input coupler interferometer , dan mereka kemudian digabungkan dalam coupler output interferometer . Panjang optik dua lengan yang tidak sama , sehingga fase sesuai dengan menunda Fig.1.1 menjadi fungsi dari panjang gelombang . relatif fase cahaya di dua port input output coupler adalah Oleh karena itu, fungsi dari panjang gelombang . Sebagai tahap keterlambatan ( d ) meningkat , siklus MZI antara lintas negara , di mana sebagian besar cahaya muncul dalam Waveguide pada sisi yang sama dengan input , dan negara bar , di mana sebagian besar bergerak cahaya untuk Waveguide di sisi lain .

Fig.1.1 Mach - Zehnder Inferometer

Sistem Komunikasi Serat Optik

Kinerja Menggunakan MZI Switching

Sachin Kumar , Indu Bala Pauria , Anoop Singhal

Kinerja Sistem Komunikasi Serat Optik Menggunakan MZI Switching

1.2 Semiconductor Optical Amplifier

Semiconductor amplifier optik adalah amplifier yang menggunakan semikonduktor untuk menyediakan media gain . desain terbaru termasuk pelapis anti - reflektif dan Waveguide miring dan jendela daerah yang dapat mengurangi refleksi akhir tatap kurang dari 0,001 % . Karena ini menciptakan hilangnya daya dari rongga yang lebih besar dari keuntungan mencegah penguat dari bertindak sebagai laser . Amplifier tersebut sering digunakan dalam sistem telekomunikasi dalam bentuk serat – pigtailed komponen , yang beroperasi pada panjang gelombang sinyal antara 0,85 ƒÊm dan 1,6 ƒÊm dan menghasilkan keuntungan hingga 30 dB . itu semikonduktor amplifier optik adalah ukuran kecil dan elektrik dipompa . Hal ini dapat berpotensi lebih murah dibandingkan EDFA dan dapat diintegrasikan dengan laser semikonduktor , modulator , dll Namun , kinerja yang masih belum 5 sebanding dengan EDFA . SOA memiliki noise yang lebih tinggi , lebih rendah gain , dan moderat ketergantungan polarisasi dan tinggi nonlinier dengan cepat sementara waktu . Ini berasal dari nanodetik pendek atau kurang seumur hidup atas negara , sehingga gain bereaksi dengan cepat terhadap perubahan pompa atau kekuatan sinyal dan Perubahan keuntungan juga karena perubahan fase yang dapat mendistorsi sinyal . Nonlinier ini menyajikan paling parah masalah bagi aplikasi komunikasi optik . Namun itu memberikan kemungkinan untuk keuntungan dalam panjang gelombang yang berbeda daerah membentuk EDFA .

1.3 Kategori switch

1.3.1 MZI Beralih

The Mach - Zehnder interferometer ( MZI ) switch berbasis

terdiri dari splitter 3 dB dan Combiner 3 dB , dihubungkan dengan dua lengan interferometer . Dengan mengubah bias efektif Indeks dari salah satu lengan , perbedaan fase di awal dari combiner dapat diubah , sehingga lampu dari port satu output yang lain . Switch ini memiliki keuntungan bahwa fase pergeseran bagian dan modus kopling bagian dipisahkan , sehingga keduanya dapat dioptimalkan secara terpisah .

Beralih berbasis Fig.1.2 MZI

Perubahan indeks bias kecil efektif dalam interferometer cukup untuk switching . Kerugiannya adalah panjangnya dan akurat perubahan indeks bias yang diperlukan untuk switching. Ketika multimode gangguan skrup yang dipekerjakan sebagai 3 dB splitter dan combiner , toleran fabrikasi dan polarisasi gelombang sensitif membimbing struktur adalah diperoleh . Sebuah sinyal data daya rendah difokuskan ke pusat masukan waveguide sedemikian rupa sehingga terbagi menjadi dua bagian yang sama di Y - junction daya splitter . Kedua balok kemudian merambat melalui dua lengan Mach - Zehnder dan bergabung kembali konstruktif pada 6 keluaran Y - junction power combiner dan merambat sepanjang waveguide output. Sebuah kontrol daya tinggi sinyal juga difokuskan ke salah satu pemandu gelombang luar untuk menghasilkan perubahan indeks bias nonlinier di Waveguide melalui nonlinear efek Kerr optik . Ini menghasilkan fase perbedaan antara dua sinyal data pada output Y junction menyebabkan mereka untuk mengganggu destruktif ketika perbedaan fasa antara mereka adalah radian TC . Di bawah ini kondisi , sinyal data yang digabungkan ke dalam mode radiasi dan output turun ke nol . Selanjutnya perangkat dapat digunakan sebagai modulator .

1.3.2 DC Beralih

Dalam switch directional coupler dua pandu gelombang yang berdekatan adalah dirancang sedemikian rupa , bahwa cahaya dapat ditransfer dari satu Waveguide ke yang lain dengan kopling . Switching diperoleh dengan benar menyesuaikan indeks bias efektif salah satu pandu gelombang . Untuk beralih hanya indeks bias kecil perubahan yang dibutuhkan .

Gambar 1.3 Directional Coupler Beralih

Untuk transfer yang baik dari cahaya , panjang kopling akurat adalah diperlukan . Karena panjang ini biasanya polarisasi dan tergantung panjang gelombang dan sangat dipengaruhi oleh fabrikasi penyimpangan ( kedalaman etch , Waveguide spasi ) , switch yang baik kinerja sulit diperoleh .

1.3.3 berbasis SOA MZI Beralih

Sebuah semikonduktor amplifier optik dapat digunakan baik untuk amplifikasi dan redaman sinyal optik , dengan memutar keuntungan dan mematikan . Properti ini dapat digunakan untuk cara sederhana namun efektif untuk beralih dengan memisahkan optik sinyal dengan 3 dB splitter , setelah sinyal ini dilemahkan di satu tangan dan diperkuat di lengan lainnya . Sejak splitter kerugian dan kerugian tambahan (misalnya serat -chip kerugian kopling ) dapat dikompensasi oleh SOA , jenis switch dapat memiliki rendah kerugian atau bahkan keuntungan dan , di samping itu, rasio yang sangat baik on-off yang mengarah ke tingkat crosstalk rendah .

Berdasarkan Gambar 1.4 SOA MZI Beralih

Kerugian yang paling penting dari sebuah saklar SOA adalah tinggi tingkat tambahan kebisingan di ? ? \ on ? \ negara akibat spontan emisi yang dihasilkan dalam SOA .

1.4 OPTSIM

Optsim adalah sistem komunikasi optik canggih paket simulasi yang dirancang untuk teknik profesional dan penelitian mutakhir dari WDM , DWDM , TDM , CATV , optik LAN , bus paralel optik , dan lainnya muncul optik sistem dalam telekomunikasi , datacom , dan aplikasi lainnya . Hal ini dapat digunakan untuk merancang sistem komunikasi optik dan mensimulasikan mereka untuk menentukan kinerja mereka mempertimbangkan berbagai parameter komponen .

100

Ara 1,5 Optsim editor grafis

Optsim dirancang untuk menggabungkan akurasi terbesar dan pemodelan kekuatan dengan kemudahan penggunaan pada kedua Windows dan UNIX platform . Optsim merupakan komunikasi optik sistem sebagai set saling berhubungan blok , dengan setiap blok mewakili komponen atau subsistem dalam komunikasi sistem . Sebagai sinyal fisik lewat di antara kompone dalam sistem komunikasi dunia nyata , ? ? \ sinyal . Data dilewatkan antara model komponen dalam simulasi Optsim .

II PUSTAKA SURVEY

2.1 All- Optical Logic oleh MZI saklar

Koji Igarashi et al . dijelaskan pemrosesan sinyal optik berdasarkan modulasi fase optik dan selanjutnya optik filtering , yang berlaku untuk 160 - Gb / s optik time-division multiplexing ( OTDM ) subsistem . ultrafast modulasi fase sinyal optik dilakukan dengan self- phase modulasi ( SPM ) dan modulasi silang - fase ( XPM ) ketika pulsa optik melewati serat optik nonlinier . demikian modulasi fase menginduksi pergeseran spektral optik sinyal . Jian Wang ET . Al . disajikan logika ultrafast gerbang AND untuk kembali - ke - nol ( CSRZ ) sinyal carrier- ditekan oleh memanfaatkan dua jenis mengalir orde kedua nonlinier dalam neonatus lithium poled berkala ( PPLN ) Waveguide .

Solusi analitis berasal di bawah tidak ada deplesi pendekatan jelas menggambarkan prinsip operasi .

Pertama , berdasarkan mengalir generasi kedua - harmonik dan generasi - perbedaan frekuensi ( CSHG / DFG ) di PPLN , sebuah semua - optik 40 Gb / s CSRZ logika gerbang AND adalah berhasil diimplementasikan dalam percobaan dan diverifikasi oleh numerik simulasi . Hal ini ditemukan bahwa pemalas dikonversi , mengambil DAN hasilnya , terus format modulasi CSRZ unchanged.Second , dengan menggunakan mengalir sum – dan generasi - perbedaan frekuensi ( CSFG / DFG ) di PPLN . oleh memodifikasi desain yang sudah ada dua masukan nano fotonik Gerbang , yang operasi didasarkan pada optik dekat lapangan ( ONF ) interaksi antara tiga titik kuantum tetangga ( Qds ) , mereka meningkatkan rasio gerbang ON / OFF hingga sekitar 9 dB . Untuk melakukan hal ini , Arash Karimkhani et al . telah menghilangkan kemungkinan langsung ONF interaksi antara input dan titik output. Kemudian , oleh menambahkan QD lain , sebagai kontrol dot kedua baik yang ada dan dimodifikasi arsitektur dua - masukan , mereka mengusulkan dua baru tiga - masukan nanophotonic gerbang schemes.one dengan interaksi ONF langsung antara input dan output titik , dan yang lain tanpa interaksi tersebut . Flip-flop adalah terdiri dari dua lintas digabungkan VCSOA elektrik dipompa inverter dan menggunakan prinsip-prinsip modulasi cross-gain , gain polarisasi anisotropi , dan keuntungan yang sangat nonlinear karakteristik untuk mencapai fungsi flip-flop . mereka menyoroti bahwa , ketika terintegrasi pada chip , jenis ini semua - optik flip-flop membuka prospek baru untuk mengimplementasikan semua - optik kenangan cepat dan sirkuit waktu regenerasi .

Jingsheng Yang et al . disajikan strategi fungsi -lock untuk semua - optik gerbang logika ( AOLG ) memanfaatkan cross- polarisasi modulasi ( CPM ) efek dalam semikonduktor amplifier optik ( SOA ) . Dengan memonitor kekuatan logika cahaya , strategi menyadari metode terkendali untuk menangkap OR dan NOR fungsi dan beralih di antara mereka . Strategi ini telah berhasil diterapkan dalam percobaan dengan 10 - Gb / s tidak - kembali-ke - nol ( NRZ ) sinyal , yang memiliki tinggi Keberhasilan -rate di atas 95 % dan memastikan rasio kepunahan yang tinggi hasil cahaya di atas 11,4 dB . Setiap langkah dalam strategi memiliki evaluasi numerik yang pasti , yang memberikan potensi pelaksanaan otomatis .

2 . 2 OTDM BY MZI SWITCHING

D. Petrantonakis , P. Zakynthinos et . al mendemonstrasikan semua - optik empat panjang gelombang modus 3R meledak regenerator , bebas dari kesalahan operasi dengan 10 - Gb / s paket data panjang variabel bahwa pameran 6 - dB variasi daya packet - to- paket . itu sirkuit dilaksanakan dengan menggunakan urutan tiga terintegrasi berbasis amplifier semikonduktor quadruple optik Array interferomentric Mach.Zehnder . T. Ohara , H. Takara et . al memberikan laporan pertama dari 160 - Gb / s optik time-division multiplexing transmisi dengan semua – channel independen modulasi dan semua -channel simultan demultiplexing . Dengan menggunakan multiplexer dan demultiplexer berdasarkan poled berkala neonatus lithium dan semikonduktor amplifier optik hibrida cahaya planar terintegrasi gelombang sirkuit , 160 - km transmisi ini berhasil demonstrated.Colja Schubert et al . menyelidiki tiga interferometric switch semua - optik yang berbasis di kayu salib fase modulasi ( XPM ) dalam semikonduktor amplifier optik ( SOA ) , laser semikonduktor amplifier di cermin lingkaran ( Slalom ) switch, interferometer Mach.Zehnder ( MZI ) switch, dan interferometer ultrafast nonlinier ( UNI ) beralih . Switching jendela dengan lebar yang berbeda diukur dalam kondisi yang sama untuk ketiga beralih konfigurasi ? e . J. M. Verdurmen disorot waktu semua – optik domain add -drop multiplexing untuk fase termodulasi OTDM sinyal untuk pertama kalinya , untuk pengetahuan kita . XIN Ming , et . al menyatakan alternatif untuk label swapping , sebuah Skema label stripping semua - optik berdasarkan SOA - MZI . itu Proses pengupasan adalah mengendalikan diri tanpa proses sinkronisasi . Hasil simulasi menunjukkan bahwa tinggi kualitas pengupasan dapat dicapai , dengan tidak lebih dari 0.09dB fluktuasi daya dan 0.05dB fluktuasi fase di kedua ditelanjangi dan tetap label . Sebuah rasio kontras kekuatan 28dB antara label dilucuti tetap dan residu, dan 30dB ratio ( SNR ) signal-to -noise dapat dicapai masing-masing. Spalter et . al . sifat transmisi dinyatakan dan kecepatan tinggi beralih teknologi disajikan untuk 160 - Gb / s OTDM sistem , yang perlu membuktikan biaya - efektif dalam point- to-point menghubungkan transmisi dan harus menawarkan rute waktu – domain kemampuan untuk menjadi kenyataan komersial . Analisis toleransi parameter menunjukkan bahwa pengupasan kinerja memburuk sedikit ketika mempertimbangkan perangkat ? ? e ketidaksempurnaan dalam praktek . Hasil simulasi multi- hop juga menunjukkan bahwa skema kami berlaku untuk OPS skala besarjaringan . Hans- Georg Weber et al . disajikan ultrahigh – speed Kinerja Sistem Komunikasi Serat Optik Menggunakan MZI Switching 101

transmisi data pada serat optik berdasarkan waktu optikdivision multiplexing ( OTDM ) teknologi transmisi . Pemrosesan sinyal optik dalam pemancar dan penerima sebagai serta persyaratan pada transmisi data kecepatan ultrahigh melalui link fiber dibahas . Akhirnya , hasil dari beberapa Percobaan OTDM - transmisi , termasuk 160 - Gb / s pengiriman melalui 4320 km , 1.28 - Tb / s transmisi lebih dari 240 km , dan 2,56 - Tb / s transmisi melalui serat penghubung 160 - km , yang dijelaskan .

2.3 Tujuan

Dalam tesis ini , penelitian dilakukan dengan tetap melihat dari tujuan sebagai berikut .

1 . Untuk mengetahui tingkat kesalahan bit dan power control dari 4 X40 Gb / s waktu optik domain sistem multiplexing menggunakanMach - Zehnder beralih .

2 . Untuk menyelidiki operasi logis optik multiplexer dan encoder menggunakan Mach - Zehnder Inferometer .

3 . Untuk mengetahui tingkat kesalahan bit dari FTTH pada 40 Gbit / s denganMach - Zehnder Switching .

2.4 Penelitian Garis Setelah mempelajari pengenalan , survei literatur dasar, kita menentukan tujuan dalam bab II .

Dalam bab III , kita menyelidiki operasi logis optik multiplexer dan encoder oleh Mach - Zhender Inferometer pada 10 Gbit / s .

Dalam bab IV , kami praktis menyelidiki dan memvalidasi bit Tingkat dan kontrol kekuatan menormalkan kekuatan Mach - Zehnder beralih pada empat saluran yang berbeda pada waktu yang berbeda bergeser pada bit rate yang sama dari 40 Gb / s .Kami akhirnya mendiskusikan kesimpulan dalam bab V dan juga pekerjaan di masa depan .

III implentation OF ENCODER OPTIK DAN

MULTIPLEXER MENGGUNAKAN MACH - ZEHNDER

INFEROMETER

Dalam bab ini perangkat logika sederhana semua - optik , yang disebutMach Zhender Inferometer terdiri dengan menggunakan Semiconductor Optical Amplifier ( SOA ) dan optik coupler . Perangkat ini digunakan untuk menghasilkan logis fungsi ( AND, XOR ) dan multiplexer dan Encoder adalah diperoleh dengan menggunakan perangkat ini di Optik Pohon Arsitektur . itu simulasi Encoder dan Multiplexer dilakukan pada tingkat 10 Gbit / s dan keduanya simulasi untuk input yang berbeda logis kombinasi . Simulasi menunjukkan bahwa perangkat ini cocok untuk beroperasi pada bit rate yang lebih tinggi dan juga untuk berbeda logis entitas .

3.1 PENDAHULUAN

Seperti kita ketahui dalam beberapa hari terakhir penelitian dalam komputasi optik meningkat dari hari ke hari dan banyak ilmuwan yang bekerja pada mereka, tetapi dalam elektronik komputasi operasi logis memainkan peran yang sangat penting karena mereka membutuhkan daya yang lebih kecil , seperti mereka adalah sirkuit digital dan dibandingkan dengan sirkuit analog , mereka sangat fleksibel . Tapi mereka memiliki kelemahan tertentu juga bahwa mereka bekerja sampai frekuensi yang terbatas , tetapi jika kita menggunakan itu logika menggunakan instrumen optik kemudian memberikan stabilitas yang lebih baik , kecepatan yang lebih baik dan switching. Dalam komputasi optik digital , sistem interkoneksi optik adalah primitif yang merupakan berbagai algoritma optik dan arsitektur . tinggi mempercepat gerbang logika semua - optik merupakan elemen kunci dalam berikutnya jaringan optik generasi dan sistem komputasi untuk melakukan fungsi pemrosesan sinyal optik , seperti semua - optik label swapping , pengakuan header, paritas memeriksa , penambahan biner dan enkripsi data. Dalam beberapa terakhir tahun , beberapa pendekatan telah diusulkan untuk mewujudkan berbagai gerbang logika baik menggunakan serat nonlinier tinggi atau semikonduktor amplifier optik ( SOA ) . The berbasis SOA perangkat memiliki potensi integrasi monolit , yang menawarkan keuntungan dari kekompakan , meningkat kehandalan dan pengurangan biaya . Hingga saat ini , berdasarkan sebagian SOA gerbang logika telah dilakukan dengan menggunakan lintas gain modulasi ( XGM ) dan modulasi silang - fase ( XPM ) , yang pasti membatasi kecepatan operasi perangkat tersebut karena dengan waktu pemulihan yang lambat pembawa intrinsik SOA.Although kecepatan operasi dapat ditingkatkan menjadi 40GB / s atau lebih tinggi dengan penggunaan daya tinggi terus menerus gelombang memegang balok [ 48 ] atau struktur interferometer yang berbeda , kompleksitas dan biaya perangkat meningkat . Permintaan kecepatan tinggi pemrosesan sinyal semua - optik telah ditimbulkan oleh arus dan jaringan optik dekat - masa depan dalam upaya untuk melepaskan node jaringan dari latency yang tidak diinginkan dan kecepatan keterbatasan yang ditetapkan oleh O / tahap konversi E / O dan untuk mencocokkan

pengolahan dan kecepatan transmisi . Dalam hal ini, peningkatan yang signifikan dalam upaya penelitian terhadap penyebaran kecepatan tinggi semua pemrosesan sinyal optik teknologi , konsep aplikasi dan demonstrasi memiliki telah menyaksikan selama beberapa tahun terakhir . semikonduktor penguat optik ( SOA ) berbasis , gerbang optik interferometric telah muncul sebagai pemrosesan sinyal fotonik main –stream unit , mengeksploitasi respon cepat mereka untuk operasi kecepatan tinggi dan mengambil keuntungan dari kemajuan yang luar biasa hibrida dan teknik integrasi monolitik untuk menawarkan kompak elemen switching . Untuk tujuan ini, elemen tunggal , kecepatan tinggi gerbang -optik telah dibuktikan sebagai perangkat terintegrasi di sejumlah laboratorium di seluruh dunia dan telah dikembangkan sebagai produk komersial terutama untuk panjang gelombang konversi dan regenerasi tujuan .

3.2 MULTIPLEXER

Multiplexer atau mux adalah perangkat yang melakukan multiplexing ; akan memilih salah satu dari banyak sinyal input analog atau digital dan output yang menjadi satu baris . Multiplexer input 2n memiliki n pilih bit , yang digunakan untuk memilih baris masukan untuk mengirim ke output . Input A Input B output

Tabel 3.1 Tabel Kebenaran 2:1 Multiplexer

3.3 ENCODER

Sebuah encoder adalah perangkat , sirkuit , transduser , program perangkat lunak dan algoritma yang mengubah informasi dari satu format , atau kode yang lain , untuk tujuan standardisasi , kecepatan , kerahasiaan , keamanan , atau menghemat ruang dengan mengecilkan ukuran . sebuah encoder dapat menjadi sebuah perangkat yang digunakan untuk mengubah sinyal ( seperti sedikit stream) atau data ke dalam kode . Kode ini menyajikan salah satu dari nomor tujuan seperti mengompresi informasi untuk transmisi atau penyimpanan, mengenkripsi atau menambahkan redudansi untuk input kode , atau menerjemahkan dari satu kode yang lain . Hal ini biasanya dilakukan dengan cara algoritma diprogram , terutama jika ada bagian digital , sementara sebagian besar encoding analog dilakukan dengan analog sirkuit .

3.4 KERJA MULTIPLEXER

Seperti yang sudah kita bahas saklar MZI untuk semua – optik logika jadi di sini kerja pohon optik menggunakan MZI berbasis switch optik . Ada sumber konstan CW berkas yang mungkin menjadi sumber laser. Sinyal cahaya yang berasal dari CWLS dapat diambil sebagai sinyal yang masuk . incoming sinyal cahaya adalah insiden pada saklar s1 pertama . Sekarang kita bisa memperoleh cahaya di cabang yang diinginkan berbeda atau sub cabang oleh menempatkan tepat dari sinyal kontrol . Sinyal kontrol juga sinyal cahaya .

Ara 3,1 logika Optical menggunakan MZI saklar

Kasus 1 : ? ? ? ? Ketika A = e0 f dan B = e0 f ?

The CW sinar yang berasal dari CWLS konstan insiden pada saklar s1 pertama . Seperti di sini A = 0,0 ? ? E , sinyal kontrol A tidak ada, itu berarti sinyal cahaya yang masuk hanya hadir pada

s1 . Sesuai prinsip beralih dibahas di atas , cahaya

muncul melalui saluran yang lebih rendah dan jatuh pada saklar s3 di C.

Berikut kontrol sinyal B tidak ada. Sebagai sinyal B yang absen sehingga

light akhirnya keluar melalui saluran yang lebih rendah dan s3

mencapai output 1 . Dalam hal ini , tidak ada cahaya hadir di lain

output port , port1 sehingga output adalah salah satu negara dan yang lain berada dalam

state nol .

Kasus 2 : ? ? ? ? Ketika A = e0 f dan B = e1 f ?

Cahaya dari sumber cahaya CW adalah insiden pada s1 . Sebagai A = .0 ? E ? ,

sinar muncul melalui saluran yang lebih rendah dan jatuh pada

s3 . Pada s3 kontrol sinyal B hadir . Dalam kehadiran

sinyal kontrol muncul melalui saluran atas dan s3

akhirnya mencapai ke port output 2 . Dalam hal ini cahaya hanya

hadir dalam output port 2 . pelabuhan Oleh karena itu keluaran menunjukkan satu negara

sementara yang lain menunjukkan nol negara.

Kasus 3 : ? ? ? ? Ketika A = e1 f dan B = e0 f ?

Cahaya dari CWLS adalah insiden pada saklar s1 pertama . Seperti di sini A

= .1 ? ? E , kontrol sinyal A hadir . Karena itu , cahaya

muncul melalui saluran atas s1and jatuh pada s2 di

O.As B = 0,0 ? ? E , tidak ada sinyal kontrol hadir di B , yang berarti

cahaya yang keluar dari saluran yang lebih rendah s2 untuk mencapai keluaran

port 3 . Jadi keluaran port 3 adalah dalam satu negara dan lain-lain berada di nol

negara.

Kasus 4 : ? ? ? ? Ketika A = e1 f dan B = e1 f ?

Cahaya dari CWLS adalah insiden pada saklar s1 pertama . Seperti di sini A

= .1 ? ? E , kontrol sinyal input A hadir . Karena itu,

cahaya muncul melalui saluran atas s1 dan jatuh pada s2

di O. Sebagai B = 0,1 ? ? e , sinyal kontrol hadir di B. Oleh karena itu

cahaya mengikuti saluran atas s2 untuk mencapai output 4 . Jadi

output port 4 adalah dalam satu negara dan yang lain adalah nol negara.

Ara 3.2 Blok Diagram Multiplexer

3.5 HASIL DAN PEMBAHASAN SIMULASI

Bagian ini tesis bercerita tentang hasil multiplexer

dan encoder menggunakan Mach - Zehnder Inferometer untuk semua - optik

logika . Proyek ini disimulasikan dalam OPTSim 4.7.1 ditentukan dalam

Modus blok yang membawa komponen yang berbeda untuk menghasilkan

sirkuit yang diperlukan yang memberikan hasil akhirnya .

3.6 SISTEM URAIAN MULTIPLEXER

Ini diberikan di bawah angka merupakan diagram skematik

semua - optik logika multiplexer oleh MZI switch. Karena mengandung

dua generator gelombang sinus yang memiliki frekuensi 10 GHz yang

bertindak sebagai pembangkit sinyal diikuti oleh Direct Modulated Laser ,

sebagai laser yang mengubah sinyal listrik menjadi sinyal cahaya dan

Output dari kedua laser diumpankan ke coupler optik yang berisi

dua port yang disebut sebagai bar pelabuhan dan lintas pelabuhan , sekarang dari setiap lengan

dari coupler diumpankan ke MZI s1 MZIs2MZIs3 .

Semikonduktor penguat optik dan akhirnya pergi ke

coupler optik sebagai coupler optik diikuti oleh semikonduktor

penguat optik disebut Mach Zehnder Beralih dan berbeda

output dari coupler optik diumpankan ke Spectrum Analyzer .

Signal Generator menghasilkan 10 sinyal GHz dalam bentuk sinusoidal

yang diumpankan ke laser DM . Direct Mode Laser blok menunjukkan

disederhanakan gelombang kontinu ( CW ) laser. Fase kebisingan adalah

diperhitungkan dengan menghasilkan sinyal generator yang

FWHM (Full Width Half Maximum ) ditentukan oleh Laser

parameter . Dalam model dianggap pusat has193.42 THz

frekuensi emisi , 1550 nm panjang gelombang , 1.650 nm

panjang gelombang , 0dBm CW Power, 1mW CW Power, Laser yang ideal

bandwith noise , lebar garis 10 FWHM dan laser acak

fase . Skrup optik , juga disebut sebagai skrup opt , adalah

perangkat terkenal digunakan cahaya langsung dari satu sumber cahaya

untuk anggota yang menerima cahaya . Sebuah coupler optik pasif

perangkat untuk percabangan atau kopling sinyal optik . umumnya ,

coupler terpusat dengan bergabung dua serat bersama-sama sehingga

bahwa cahaya dapat melewati dari unit pengirim ke dua

penerima , atau yang lain itu dapat dilakukan dengan menyandingkan dua

" receiver " serat yang kemudian akan selaras dan diposisikan sehingga

sebagai untuk menghadapi " pengirim " serat .

Semiconductor amplifier optik adalah amplifier yang menggunakan

semikonduktor untuk menyediakan media gain . itu

semikonduktor amplifier optik adalah ukuran kecil dan

elektrik dipompa .

SOA memiliki noise yang lebih tinggi , keuntungan yang lebih rendah , polarisasi moderat

ketergantungan dan nonlinier tinggi dengan cepat sementara waktu .

Ini berasal dari nanodetik pendek atau kurang negara atas

seumur hidup , sehingga gain bereaksi cepat terhadap perubahan pompa atau

kekuatan sinyal dan perubahan keuntungan juga menimbulkan fase

perubahan yang dapat mengganggu sinyal .

Kinerja Sistem Komunikasi Serat Optik Menggunakan MZI Switching

103

Ara 3.3 Skema Diagram Multiplexer ( A = 0,1 ? ? E , B = 0,0 ? ? E )

Ara 3.4 Panjang gelombang spektrum A = 0,1 ? ? E & B = 0,0 ? ? E

Diagram di atas menunjukkan spektrum panjang gelombang

diperlukan logika pada output port 1 . Sebagai spektrum bahwa baik

sinyal input dan sinyal kontrol memiliki berbeda

panjang gelombang jadi kita gunakan untuk sinyal kontrol adalah 1550 um

sedangkan sinyal yang masuk terdiri dari panjang gelombang 1650 um

sehingga memiliki amplitudo maksimum pada panjang gelombang kontrol

sinyal .

Kasus 1 : ? ? ? ? ? ? Ketika A = f1 f , B = e0 f , & EN = e1 f ?

Dalam diagram skematik ini encoder tiga gelombang sinus

generator digunakan untuk menghasilkan pulsa sinusoidal yang secara langsung

diumpankan ke laser termodulasi langsung yang bekerja di berbagai

panjang gelombang sinyal masukan tertentu , sebagai encoder ini memiliki

tiga sinyal input dan mengaktifkan sinyal pada panjang gelombang yang berbeda

dari sinyal input dan mengaktifkan sinyal ini diumpankan langsung ke

lengan input dari coupler dari MZI beralih dengan beam splitter

dan memberikan logika yang diperlukan .

Skema diagram dari encoder ( A = 0,1 ? ? E , B = .1 ? ? E )

Panjang gelombang spektrum A = 0,1 ? ? E & B = .1 ? ? E

Kasus 2 : ? ? ? ? ? ? Ketika A = e1 f , B = e1 f & EN = e1 f ?

Dalam skema ini sebagai diagram terakhir mewakili dua sinusoidal

Generator 10 GHz diikuti oleh laser Direct Mode yang

mengubah sinyal listrik menjadi sinyal optik atau sinyal cahaya

dan output dari laser langsung dimasukkan ke lengan input

coupler yang melewati sinyal pada bar pelabuhan sebagai tergantung

pada sinyal kontrol . Di sini sinyal kontrol dimasukkan ke dalam

sirkuit di tingkat ketiga saklar MZI karena terdiri dari

dua semikonduktor amplifier optik di pelabuhan kedua

coupler optik pada input dan hal yang sama diikuti di

output dari saklar.

Jadi di sini dalam sinyal kontrol sirkuit ini diterapkan untuk semua yang

masukan dari encoder tetapi menurut prinsip MZI

menonaktifkan input diterima di pelabuhan bar coupler ketika

sinyal kontrol hadir sehingga kami menerapkan dua kontinu

sinyal pada masukan dari kedua laser sehingga pada output tahap pertama adalah

diterima di pelabuhan bar dari coupler optik 2 sesuai dengan

sehingga output dari coupler optik diumpankan ke input dari

coupler optik 3 dan pada saat yang sama masukan ketiga juga FBI untuk

input dari coupler optik 3 sekarang lagi sama

fenomena hadir sebagai sinyal kontrol menjadi output dari

coupler optik 2 dan berkesinambungan sinyal gelombang diperlakukan sebagai masukan

output sehingga sama sama dari coupler optik diumpankan ke

input coupler optik 5 dan juga sinyal kontinyu yang sama

diumpankan ke coupler optik 4 dan sekarang output dari optik

coupler 3 & 5 diproses dengan benar .

Skema diagram dari encoder ( A = 0,1 ? ? E , B = 0,0 ? ? E , EN = ? ? E1 ? E )

Panjang gelombang spektrum ( A = 0,1 ? ? E & B = 0,0 ? ? E , EN = ? ? E1 ? E )

Sekarang output dari coupler optik 4 & 6 diumpankan ke salah satu

masukan lengan 7,9 serat optik , 11 , 13 dan kemudian output

dari ini coupler optik yang diperlukan adalah pergi ke optik

coupler 8 , 10 , 12 , 14 sampai lewat dengan Semiconductor yang

penguat optik ( SOA ) . Pada kedua port dari optik

spektrum analyzer coupler dihubungkan untuk mengukur

spektrum panjang gelombang yang lewat melalui tepat

channel seperti yang kita lihat sebelumnya jika kita menerapkan masukan pada kedua ujung

coupler salah satu dari terus menerus dan lainnya adalah sinyal kontrol

memiliki panjang gelombang yang berbeda dari gelombang kontinu

sinyal maka output diterima pada satu port coupler

sehingga dengan cara itu MZI beralih bekerja sebagai inverter logis sehingga

EN sini sama seperti bekerja sebagai inverter begitu berbeda spektrum

diterima di output dari coupler optik tapi benar

cara output diterima di spektrum analyzer 10 dan

ditampilkan di ara 2.10 yang menunjukkan spektrum panjang gelombang

sinyal yang diterima dari semua logika optik encoder dalam bentuk

.1 ? E .

International Journal Soft Computing dan Teknik ( IJSCE )

ISSN : 2231-2307 , Volume - 2 , Edisi - 3 , Juli 2012

104

Ara 3.5 Skema diagram dari encoder ( A = 0,1 ? ? E , B = .1 ? ? E , EN =

.1 ? ? E )

Gambar 3.6 Panjang gelombang spektrum A = 0,1 ? ? E & B = .1 ? ? E , EN = 0,1 ? ? E

3.7 KESIMPULAN

Kami telah disimulasikan logika berbasis Multiplexer semua - optik dan

Encoder menggunakan MZ Inferometer . Fungsi logika yang berbeda

dapat direalisasikan hanya dengan menyesuaikan dua komponen yaitu

multiplexer dan encoder. Metode simulasi memiliki

potensi untuk beroperasi diatas 40 GB / s .

IV OPTICALTIME DIVISION MULTIPLEXING MENGGUNAKAN

MZI SWITCHING

Banyak aliran data yang lebih rendah kecepatan dapat di-multiplexing ke

satu kecepatan tinggi aliran dengan cara pembagian waktu Optical

multiplexing ( OTDM ) , sehingga setiap saluran masukan mentransmisikan

data dalam slot waktu yang ditetapkan . Tugas ini dilakukan

oleh sebuah saklar multiplexer cepat ( mux ) . The routing yang berbeda

data stream pada akhir link TDM dilakukan oleh

beralih demultiplexer ( demux ) dan demultiplexer ini

bekerja menggunakan MZI beralih karena terdiri semikonduktor

penguat optik ( SOA ) dan coupler optik . Dalam bab ini

empat saluran OTDM disimulasikan pada 40 Gbit / s dan selanjutnya adalah

meneliti dampak dari kekuatan sinyal , lebar pulsa dan

mengontrol kekuatan sinyal pada BER .

4.1 PENDAHULUAN

Kapasitas transmisi jaringan optik bisa

diperpanjang dengan cara yang sederhana dengan memasang serat tambahan (spasi

division multiplexing atau SDM ) . Karena ini sangat mahal ,

metode telah dikembangkan untuk penggunaan yang lebih efisien dari

bandwidth yang tersedia dalam jaringan serat yang ada . A pertama

solusi adalah untuk meningkatkan bit rate dalam jaringan , yang

membutuhkan elektronik kecepatan tinggi di node jaringan .

Interleaving dapat dilakukan pada bit - by- bit dasar , seperti ,

atau pada packet - by- packet basis . Sebagai kecepatan data menjadi lebih tinggi

dan lebih tinggi , menjadi lebih sulit untuk bagian-bagian elektronik

( switch ) dalam sistem untuk menangani data dengan benar . A. Cheng

et al . Disajikan 40 Gb / s demultiplexing OTDM menggunakan

semua - optik merdu delay line dan elektro - penyerapan

modulator . Penundaan serat optik terus menerus untuk saluran

Temukan direalisasikan menggunakan empat - gelombang - pencampuran dan panjang gelombang

tergantung delay kelompok . Ken Morito et al . disajikan seragam

kekuatan output dan rasio kepunahan tinggi untuk Mach - Zehnder

Jenis interferometer semua switch optik dengan asimetris

amplifier bias dan shifter fase ditemukan di dinamis

analisis untuk kontrol pulsa sempit dan beralih dioptimalkan

windows. Kinerja pembagian waktu optik

multiplexing ( OTDM ) sistem dibatasi oleh kompleks

kombinasi suara . Dalam makalah ini kami menyajikan teoritis

kerangka kerja untuk penerima optik dalam sistem OTDM berdasarkan

fungsi generasi saat . Jianfeng Zhang et . al .

disajikan model receiver yang diusulkan menunjukkan lebih

akurat dalam memprediksi tingkat kesalahan bit ( BER ) kinerja

dari yang bekas [ 47 ] Masalah ini dapat diatasi dengan .

routing data melalui domain optik, yang dilambangkan

waktu sebagai optik division multiplexing ( OTDM ) . Kecepatan

hari ini sistem OTDM eksperimental dalam urutan

10 Gb / s ( saluran tunggal ) , dan sebagian besar dibatasi oleh kecepatan

unsur-unsur non - linear dan pengaruh efek fisik

seperti dispersi kromatik pada pulsa optik dalam

serat dipekerjakan . Mach - Zehnder interferometer dengan

terintegrasi SOA ( SOA - MZI ) sangat menarik sebagai

kecepatan tinggi gerbang optik . Mereka memiliki energi beralih rendah ,

kekompakan tinggi dan stabilitas , serta potensi

integrasi optik lebih lanjut. Dalam bab ini kita simulasi empat

saluran saluran OTDM pada kecepatan 40 Gbit / s untuk BER dengan

lebar pulsa , Pengendalian Sinyal Daya .

4.2 WAKTU DIVISION MULTIPLEXING

TDM adalah proses digital yang memungkinkan beberapa sambungan ke

berbagi bandwidth tinggi link . Alih-alih berbagi a

sebagian dari bandwidth seperti pada FDM , waktu bersama . masing-masing

koneksi menempati sebagian waktu di link . TDM adalah

Teknik multiplexing digital untuk menggabungkan beberapa tingkat rendah

saluran menjadi satu tingkat tinggi satu . Pada TDM , data rate dari

link n kali lebih cepat , dan durasi unit n kali lebih pendek .

4.3 OPTIK WAKTU DIVISION MULTIPLEXING

Multiplexing elektronik pada kecepatan tersebut tetap sulit

dan menyajikan pembatasan pada pemanfaatan bandwidth

link serat single-mode . Strategi alternatif untuk meningkatkan

bit rate sistem serat optik digital di luar

kemampuan bandwidth dari drive elektronik yang dikenal sebagai

waktu optik division multiplexing ( OTDM ) . Di mulai dari

pengiriman data serat optik saluran digital listrik

sinyal elektrik telah sampai multiplexing untuk maksimal yang

Data agregat tingkat menyusul hirarki data yang telah ditetapkan .

Sinyal listrik agregat ini dikonversi

elektro - optik ke dalam domain optik hanya untuk

transmisi . Untuk demultiplexing , yang ditransmisikan optik

sinyal diubah ke dalam domain listrik dan

demultiplexed dalam domain listrik . Prinsip ini

Teknik adalah untuk memperpanjang waktu division multiplexing dengan optik

menggabungkan sejumlah kecepatan yang lebih rendah baseband elektronik

saluran digital . Gambar menunjukkan multiplexing optik dan

Rasio demultiplexing adalah 1:4, dengan tingkat saluran baseband

bit rate yang dibutuhkan .

Sistem dapat disebut sebagai sistem OTDM empat channel .

Empat pemancar pada gambar didorong oleh umum 40

Jam GHz menggunakan kuartal bit jangka waktu penundaan . modus Dikunci

Kinerja Sistem Komunikasi Serat Optik Menggunakan MZI Switching

105

sumber laser semikonduktor yang diproduksi pendek optik

pulsa yang digunakan pada pemancar untuk memberikan tugas rendah

pulsa siklus stream multiplexing untuk waktu berikutnya . data

dikodekan ke sungai pulsa ini menggunakan optik terintegrasi

MZ modulator yang memberi RZ output pemancar di 10

Gbit / s . perangkat IO ini digunakan untuk menghilangkan laser

kicauan akan menghasilkan dispersi dari pulsa ditransmisikan sebagai

mereka disebarkan dalam mode serat tunggal , sehingga membatasi

jarak transmisi dicapai .

Empat 40 Gbit / s sinyal data digabungkan dalam pasif

power combiner optik tapi , pada prinsipnya , sebuah switching yang aktif

elemen dapat dimanfaatkan . Meskipun empat sumber optik

bekerja , mereka semua dipancarkan pada panjang gelombang optik yang sama

dalam toleransi ? ? } 0,2 nm dan karenanya 40 Gbit / s Data

stream sedikit disisipkan untuk menghasilkan 160 Gbit / s baseband

komponen dalam demultiplexer yang terdiri dari dua tingkat .

Sekali lagi perangkat Waveguide IO digunakan untuk memberikan switching

berfungsi pada setiap tingkat . Pada tingkat pertama saklar IO didorong

oleh sinusoid pada 80 GHz untuk demultiplex yang masuk 160

Gbit / s stream ke 80 Gbit / s sinyal . Oleh karena itu tunggal panjang gelombang

160 Gbit / s transmisi optik diperoleh dengan elektronik

yang hanya dibutuhkan bandwidth maksimum sekitar 25 GHz ,

sebagai kembali ke nol pulsa bekerja .

4.4 DEMUX OPERASI MENGGUNAKAN MZI - SOA SWITCH

The MZI - SOA beralih semua optik ditunjukkan pada Gambar 5.4

terdiri dari dua simetris 2x2 multimode interferometer

( MMI ) splitter untuk membagi dan menggabungkan pulsa data, dua

skrup untuk memperkenalkan pulsa kontrol, dua SOA ? es untuk

memberikan pergeseran fasa , dan shifter fase ( PS ) untuk menyesuaikan

mengimbangi fase . Sinyal data diinjeksikan dari port input

diarahkan ke lintas pelabuhan atau port bar tergantung pada

perbedaan fasa antara dua es SOA ? . Dengan menyuntikkan

control pulsa 2 dengan penundaan waktu tertentu dan energi yang tepat

perbedaan terhadap pulsa kontrol 1 , perubahan fase lambat

terkait dengan pemulihan gain lambat dalam dua SOA ? es yang

benar-benar dihilangkan . Hal ini menimbulkan beralih pendek

windows. Dengan menyesuaikan kali suntikan dua control

pulsa , salah satu sinyal pulsa data yang multiplexing dapat

turun ke port bar dan sinyal lain dapat

ditransmisikan ke lintas negara . Berikut kontra menyebarkan data yang

dan pulsa kontrol diasumsikan .

Ara 4.1 DEMUX menggunakan MZI saklar

4.5 SETUP SIMULASI UNTUK OTDM

Setup sistem tertentu dari OTDM ditunjukkan pada gambar

( 5.5 ) . Komponen yang digunakan dalam gambar ( 5.5 ) dipilih dari

Optsim Ver.4.7.0 palet komponen perpustakaan dan ditempatkan sesuai

persyaratan di bidang desain editor Optsim . kemudian

berbagai parameter simulasi ditetapkan . pemancar

terdiri dari urutan biner pseudo- acak atau PRBS

Generator , modus terkunci dioda laser , sebuah generator listrik ,

empat kali pergeseran blok , sebuah MUX optik dan optik

menormalkan . Beberapa saluran dari MLLD adalah RZ

dimodulasi dengan pola PRBS yang berbeda . The PRBS block

menghasilkan beberapa output pola , masing-masing berbeda dari

lainnya dan pada bit rate yang sama . Semua saluran dari MLLD berada di

panjang gelombang yang sama 1650nm dan kekuatan yang sama . sebelum

menjadi multiplexing bersama-sama masing-masing saluran konsekuen adalah

tertunda oleh 1/4 dari waktu window berturut-turut . Jumlah kekuatan

semua saluran dikendalikan oleh menormalkan optik, yang

menentukan output daya rata-rata sinyal OTDM sebelum

propagasi lebih panjang serat . Para perjalanan sinyal OTDM

melalui serat optik dari 100 km panjang dan maka itu adalah

de - multiplexing pada akhir penerima . Penerima terdiri dari

empat identik beralih SMZ ( tapi dengan penundaan waktu yang berbeda ) ,

masing-masing terdiri dari generator denyut nadi kereta api ( dengan pengulangan yang sama

Tingkat sebagai pemancar ) , normalizer optik blok , splitter pulsa

dan dua blok waktu tunda dan SMZ beralih dengan dua keluaran

port . The BER meter terhubung pada kedua output dan

tercermin pelabuhan untuk mendapatkan hasil . Semua saklar SMZ adalah

terhubung pada output dari fiber.Schematic nonlinear

Diagram OTDM menggunakan MZI beralih

4.6 HASIL DAN PEMBAHASAN

Sinkronisasi antara pemancar dan penerima di OTDM

merupakan isu penting untuk kinerja optimal dari sistem . dalam hal ini

kertas, pemancar dan penerima telah disinkronkan

dengan penambahan delay optik dalam sinyal kontrol . itu

delay optik bervariasi sebagai multiple integer dari 1/4 dari

lebar pulsa dalam sebuah diharapkan terikat. Pola yang

muncul dari variasi tersebut menentukan optimum optik

delay yang dibutuhkan untuk setiap saluran . Pengaruh kebisingan dan

distorsi sangat terkenal di transmisi digital . kebisingan

menyebabkan kesalahan bit di gerbang keputusan penerima dan

distorsi menyebabkan perubahan pada bentuk pulsa sehingga antar

gangguan simbol ( ISI ) , yang juga menghasilkan kesalahan bit .

Parameter utama selain bandwidth, yang

ciri link optik digital , adalah BER . Jadi efek

kekuatan sinyal ( Psignal ) , sinyal kontrol listrik ( Pcontrol ) , dan

lebar pulsa pada BER diselidiki . Gambar menunjukkan variasi

BER dengan perubahan dalam kekuatan sinyal . Seperti disebutkan sebelumnya

normalizer optik mengontrol daya keluaran rata-rata

Sinyal multiplexing .

BER untuk kanal 1 adalah di kisaran 10-21.10-28 untuk

Psignal nilai 5 dan 10 dBm , masing-masing.

Ara 4.2 BER versus daya sinyal input dengan dispersi

Jadi teramati bahwa dengan peningkatan kekuatan sinyal

( Psignal ) BER mengalami perbaikan .

Demikian pula untuk saluran 2 dan 4 variasi ini berada di kisaran

10-22.10-25 dan 10-22.10-25 untuk nilai Psignal dari 5 dan 10

International Journal Soft Computing dan Teknik ( IJSCE )

ISSN : 2231-2307 , Volume - 2 , Edisi - 3 , Juli 2012

106

dBm , masing-masing. Sangat menarik untuk dicatat bahwa BER untuk

saluran 2 dan 3 adalah sama untuk semua nilai Psignal . BER dari

penerima optik berbanding terbalik dengan SNR , yang di

menghidupkan tergantung pada daya optik sinyal . Jadi BER

menurun dengan peningkatan kekuatan sinyal .

Ara 4.2 BER vs lebar pulsa dengan dispersi

Selanjutnya pada Gambar . 5.8 efek perubahan lebar pulsa pada BER

diselidiki . Lebar pulsa dari sinyal input bervariasi

dalam batas-batas 5e - 12.12e - 12 m dan variasi dalam BER

diamati . Seperti yang terlihat pada gambar untuk saluran 1 BER pada

5e - 12m adalah 10-140 dan dengan peningkatan denyut nadi lebar itu

menurun menjadi 10-27 untuk lebar pulsa 12e - 12 m . sekali lagi

ada tumpang tindih dalam kurva untuk saluran 2dan 3 dan

variasi untuk BER adalah 10-140 ke 10-167 untuk

Variasi tersebut di atas lebar pulsa . Untuk channel 4 yang

nilai BER bervariasi 10-140 hingga 10-162 untuk

yang disebutkan di atas variasi lebar pulsa . hasil

menunjukkan peningkatan kinerja penerima dengan

peningkatan lebar pulsa . Peningkatan ini dapat dikaitkan dengan

pengurangan lebar pulsa distorsi Menampilkan signifikan

degradasi kinerja penerima ketika sinyal kontrol

listrik meningkat secara bertahap melampaui 22 dBm .. Jadi dalam kasus

channel 1BER pada sinyal kontrol 22dBm adalah 10-35 dan meningkat

untuk 10-4 pada 26 dBm . Saluran 2 dan 3 sekali lagi menunjukkan

pola BER identik dan variasi dalam

Ara 4.3 BER vs sinyal kontrol listrik dengan dispersi

Rentang 10-35.10-4 pada 22 dan 26 dbm , masing-masing. itu

variasi BER untuk kanal 4 adalah di kisaran 10-35.10-2

untuk variasi yang disebutkan di atas dalam kekuatan sinyal kontrol. itu

dipahami bahwa prinsip operasi dari sebuah switch adalah MZI

berdasarkan interferensi antara sinyal melewati dua

kaki dari MZI . Sinyal kontrol mempengaruhi perubahan

indeks bias bahan semi - konduktor . Perubahan

indeks bias pada gilirannya memperkenalkan pergeseran fasa pada input

sinyal . Ini menunjukkan efek Pcontrol pada BER tanpa

dispersi untuk semua channels.the dua sinyal gangguan pada

output dan output yang dihasilkan tergantung pada mereka

pergeseran fase relatif . Dengan demikian , sinyal dapat mengganggu baik

konstruktif atau destruktif . Dari grafik jelaslah

bahwa peningkatan sinyal kontrol melampaui 22 dB m memperkenalkan

pergeseran fasa , yang menurunkan kinerja penerima dan

BER terus meningkat dengan peningkatan sinyal kontrol

kekuasaan. Gambar . 5.12 menggambarkan diagram mata untuk channel1 , di

Nilai Pcontrol dari 22 dan 26 , masing-masing. ada

degradasi di tingkat keputusan

nilai offset dari 1,5 X 10-5 sampai 7 X 10-6 dengan peningkatan

Nilai Pcontrol 22-26 . Pengamatan ini mendukung

kesimpulan yang ditarik dari ayat-ayat BER sinyal Pcontrol .

Ara 4.4 Diagram Eye

4.6 KESIMPULAN

A 160 - Gb / s transmisi OTDM dengan semua - channel

modulasi dan semua -channel demultiplexing simultan memiliki

berhasil disimulasikan untuk pertama kalinya . The MUX dan

DEMUX penggunaan MZI beralih ketat menjaga waktu tunda

antara saluran yang berdekatan dan suhu tinggi - tawaran

stabilitas karena mereka hibrida terintegrasi pada MZI saklar ;

mereka akan , oleh karena itu, menjadi kunci untuk transmisi OTDM masa depan

sistem .

V. KESIMPULAN DAN ASPEK MASA DEPAN

5.1 KESIMPULAN

Dalam tesis ini skema menghasilkan logika optik

dilaksanakan oleh MZ Inferometer seperti yang dibahas dalam bab 2 kaleng

digunakan untuk tujuan yang berbeda . Skema ini dapat dengan mudah dan

berhasil diperpanjang dan dilaksanakan untuk lebih tinggi

jumlah masukan angka oleh penggabungan yang tepat dari MZI berdasarkan

switch optik , ekstensi vertikal dan horizontal pohon

dan oleh seleksi cabang sesuai. Sekali lagi seluruh operasi

paralel di alam , yaitu hasil usaha yang berbeda

antara data yang diperoleh pada suatu waktu . Di sini kita bisa

melaksanakan beberapa instruksi beberapa tipe data

operasi dengan baik . Operasi aritmatika dapat dilakukan

sini antara dua data yang besar berbentuk . Salah satu yang diusulkan

bit skema perbandingan digital juga berhasil memanfaatkan

bahan non - linear berbasis struktur pohon untuk operasi . sekarang

Penting untuk dicatat bahwa diskusi di atas didasarkan pada

model sederhana . Dalam simulasi ini beberapa parameter berjalan harus

dipertimbangkan sehingga sifat dispersi , polarisasi

serat , nilai-nilai yang telah ditentukan dari intensitas / panjang gelombang

sinar laser untuk kontrol dan sinyal masuk , pengenalan

filter , kerugian intensitas karena splitter balok / fiberskrup dll Seperti dalam tesis ini panjang gelombang gelombang terus menerus sinar laser adalah 1550 ƒÊm dan sinyal berdenyut panjang gelombang dari 1650 ƒÊm dapat digunakan sebagai masuk dan kontrol sinyal , masing-masing. Intensitas kerugian akibat skrup dan splitter dalam tahap interkoneksi tidak mungkin menciptakan banyak kesulitan dalam memproduksi bit optik yang diinginkan pada output sebagai keseluruhan sistem adalah salah satu digital dan output tergantung pada keberadaan dan ketiadaan cahaya . di Kinerja Sistem Komunikasi Serat Optik Menggunakan MZI Switching

107

skrup tahap interkoneksi serat dapat digunakan sebagai pengganti splitter balok . Empat channel 4 X 40 Gb / s sistem OTDM ( semua saluran ) dengan modulator Mach - Zehnder , MZI switching dan panjang serat dari 100 Km , telah eksperimental dan berhasil diverifikasi . Hasil penelitian mengungkapkan bahwa BER menurun dengan meningkatnya kekuatan sinyal dan peningkatan lebar pulsa . Seperti dalam tesis BER ini meningkat dengan peningkatan kekuatan sinyal kontrol dengan dispersi dalam serat mode tunggal . Hal ini juga disimpulkan bahwa kinerja sistem OTDM dapat

ditingkatkan dengan menggunakan dispersi kompensasi serat. 5.2 ASPEK FUTURE logika All- optik adalah penelitian terbaru di bidang komputasi optik sebagai skema ini juga memberikan ide memori optik jika kita merancang sebuah flip-flop optik yang menyimpan data sebagai pulsa optik . Seperti FTTH memiliki banyak keuntungan atas semua teknik transmisi sehingga , Penyedia bisa menggunakan ATM , SONET , Ethernet atau Analog termodulasi pembawa RF sebagai teknologi lapisan data link mereka . Karena semua pengguna dilayani oleh splitter yang sama . Combiner pada PON tepi jalan ( dan oleh Remote Node yang sama dalam sebuah arsitektur Bintang Aktif ) harus dilayani oleh teknologi lapisan data - link yang sama . Infrastruktur FTTH yang berteknologi dan kompetitif netral ; di mana penyedia suara , video dan layanan data dapat memilih dan menggunakan teknologi pilihan mereka untuk mendukung layanan mereka berencana untuk menawarkan . FTTH juga menyediakan layanan tambahan di atasnya seperti UWB ( Ultra wide band ) , WCDMA , Radio atas serat , begitu banyak layanan lain seperti jaringan akan menggunakan jaringan FTTH sebagai interface untuk jaringan akses . Fokus telah dimasukkan pada sistem PON masa depan-bukti memiliki gigabit simetri dalam bandwidth antara downlinks atas dan bawah. Telah terbukti bahwa OCDMA mampu memberikan gigabit - atau bahkan multi- gigabit - per - detik untuk setiap pengguna baik di up- dan downlinks , dan OCDMA over WDM PON bisa menjadi salah satu arsitektur yang paling menjanjikan yang dapat menembus yang terakhir / pertama mil bottleneck .

REFERENSI

[ 1 ] Jitendra Nath Roy , ? ? \ Mach - Zehnder interferometer berbasis arsitektur pohon untuk semua - optik logika dan operasi aritmatika . , Optik Int Cahaya Elektron Opt . (2009) .

[ 2 ] Koji Igarashi dan Kazuro Kikuchi , ? ? \ Optical Signal Processing oleh Phase Modulation dan selanjutnya spektral Filtering Bertujuan Aplikasi untuk ultrafast Sistem Komunikasi Optik . , IEEE jurnal topik yang dipilih dalam elektronik kuantum , Vol . 14 , No 3 , Mei / Juni .

[ 3 ] K. Uchiyama , H. Takara , K. Mori , T. Morioka , ? \ 160 Gbit / s all- optical time-division demultiplexing memanfaatkan dimodifikasi multiple- output yang OTDM demultiplexer ( MOXIC ) . , Electron . Lett . 38 ( 2002) 1190,1191 .

[ 4 ] I. Shake, H. Takara , I. Ogawa , T. Kitoh , M. Okamoto , K. Magari , T. Ohara , S. Kawanishi , 0,160 - Gbit / s channel penuh optical time-division de- multiplexer berdasarkan SOA -array terpadu PLC dan aplikasi untuk OTDM percobaan transmisi . , IEICE Trans . Commun . 53 ( 1 ) ( 2005 ) 20,2096 .

[ 5 ] H. Le - Minh , Z. Ghassemlooy , W.P. Ng , ? ? \ Penindasan crosstalk dalam simetris Mach.Zehnder semua - optik ( SMZ ) beralih dengan menggunakan pulsa kontrol dengan kekuatan yang tidak sama , Prosiding Simposium Internasional tentang Telekomunikasi 2005 ( IST 2005 ) . , Vol . 1 , Shiraz , Iran 2005 , hlm 265,268 .

[ 6 ] M. Heid , S. Spalter , G. Mohs , A. Farbert , W. Vogt , H. Melchior , ? \ 160 - Gb / s demultiplexing didasarkan pada interferometer Mach.Zehnder monolithically terintegrasi , Proceedings of the European konferensi Optical Komunikasi (ECOC 2001) " , Amsterdam , Belanda , September 30.October 4 , 2001.

[ 7 ] Haijiang Zhang , Pengyue Wen , dan Sadik Esener , . Cascadable inverter semua - optik

berdasarkan vertikal - rongga semikonduktor amplifier optik nonlinier . , Opt . Lett . 32 , 1884-1886 ( 2007) .

[ 8 ] Yanming Feng , Xiaofan Zhao , Li Wang , dan Caiyun Lou , ? ? \ Kinerja tinggi semua - optik OR / NOR gerbang logika dalam semikonduktor amplifier optik tunggal dengan keterlambatan gangguan filtering , Appl .. , Opt . 48 , 2638-2641 (2009) .

? ? [ 9 ] Jitendra Nath Roy dan Dilip Kumar Gayen , \ Terpadu logika semua - optik dan arithmeticoperations dengan bantuan perangkat interferometer berbasis Toad - pendekatan alternatif MF Lane , DZ Chen , dan D. Kokkinos , Mengelola Fiber Connections di NGN dan Aplikasi , di Konferensi Fiber Optic Nasional . , OSA Teknis Digest Series ( CD ) ( Optical Society of America , 2007) , kertas NThA1 .

[ 10 ] Petrantonakis , P. Zakynthinos , D.Apostolopoulos , A.Poustie , G. Maxwell , dan H.

Avramopoulos , . All- Optical Empat - Wavelength Burst Mode Regenerasi Menggunakan Integrated Quad SOA - MZI Arrays . , IEEE Photonics TEKNOLOGI LETTERS , VOL . 20 , NO . 23 , 1 Desember 2008 .

[ 11 ] Colja Schubert , Jorn Berger , Stefan Diez , Hans Jurgen Ehrke , Reinhold Ludwig , Uwe Feiste , Carsten Schmidt , Hans G. Weber , Gueorgui Toptchiyski , Sebastian Randel , dan Klaus Petermann , ? ? \ Erbandingan Interferometric Switches All- Optical untuk Demultiplexing Aplikasi di Internet Kecepatan Tinggi OTDM Systems. , JOURNAL OF TECHNOLOGY Lightwave , VOL . 20 , NO . 4 , April 2002 .

[ 12 ] K. Kitayama , T. Kuri , JJ Vegas Olmos , dan H. Toda , ? ? \ Fiber - Wireless Networks dan Radio - over - Fiber Teknik , di Konferensi Laser dan Electro-Optics/Quantum Elektronika dan Laser Science Conference dan fotonik Aplikasi Sistem Teknologi . , OSA Teknis Digest ( CD ) ( Optical Society of America , 2008) , kertas CThR4 .

[ 13 ] R. Llorente , T. Alves , M. Morant , M. Beltran , J. Perez , A. Cartaxo , dan J. Marti , . Distribusi Optical dari OFDM dan Impulse - Radio UWB di FTTH Networks . , Di National 74 Fiber Optic Engineers Conference , OSA Teknis Digest ( CD ) ( Optical Society of America , 2008) , kertas JWA109 .

[ 14 ] Kuniharu Himeno , Shoichiro Matsuo , Ning Guan , dan Akira Wada , Low- Bending - Rugi Single- Mode? ? \ Serat untuk Fiber- to-the -Home . , J. Cahaya gelombang Technol . 23 , 3494 - (2005)

[ 15 ] D. Iazikov , C. Greiner , dan TW Mossberg , ? ? \ Filter Apodizable Terpadu untuk Rambut Kasar WDM dan FTTH -Type Aplikasi . , J. Cahaya gelombang Technol . 22 , 1402 - ( 2004)

[ 16 ] MF Lane , DZ Chen , dan D. Kokkinos , . Mengelola Fiber Connections di NGN dan Aplikasi . , Di Fiber Optic Nasional Engineers Conference , OSA Teknis Digest Series ( CD ) ( Optical Society of America , 2007) , kertas NThA1 .

0121: Anto Satriyo Nugroho dkk. TI-15

PENGEMBANGAN SISTEM COMPUTER AIDED DIAGNOSIS BERBASIS FREE OPEN SOURCE SOFTWARE

Anto Satriyo Nugroho'*, Made Gunawan', Ismail Ekoprayitno Rozi2, Vitria Pragesjvara',

Miranti Jatnia Riski', Desiani', Nuke Dewi Kania, Theda Lukito, Agung Riyadi, Yuni Arti, Ninon Nurul Faiza

'Pusat Teknologi Informasi & Komunikasi, Badan Pengkajian & Penerapan Teknologi

Gedung Teknologi 3 Lantai 3 Puspiptek Serpong, Tangerang 15314

Tel. (021) 75791260 Ext.236 Fax. (021) 75791248

2Divisi Bioinformatika, Lembaga Biologi Molekuler Eijkman

Jl. Diponegoro 69, Jakarta 10340

Tel. (021) 3917131 Fax. (021) 3147982

*e-Mail: anto.satriyo@bppt.go.id, asnugroho@gmail.com

Disajikan 29-30 Nop 2012

ABSTRAK

Malaria merupakan penyakit tropis utama yang hingga kini masih banyak ditemukan di berbagai daerah di Indonesia. Penyakit Malaria disebabkan oleh infeksi protozoa pada darah yang disebut plasmodia. Parasit ini terdiri dari 4 spesies: Plas¬modium vivax, Plasmodium falciparum, Plasmodium ovale dan Plasmodium malariae. Penyakit Malaria diindikasikan dari berbagai gejala seperti menggigil, demam tinggi, sakit kepala, anemia dan pembesaran Limfa. Di Indonesia diperkirakan jumlah penderita Malaria mencapai sekitar 15 juta, dengan angka kematian 30 ribu per tahun. Angka ini mengindikasikan bahwa Malaria merupakan salah satu penyakit tropis di Indonesia yang harus mendapatkan penanganan serius dan cepat. Salah satu faktor penting dalam penanganan kasus Malaria adalah diagnosa dini Malaria yang akurat. Badan Pengkajian Penerapan Teknologi telah bekerja sama dengan Lembaga Biologi Molekuler Eijkman mengembangkan sistem Computer Aided Diagnosis berbasis teknik pengolahan citra yang mampu membantu proses diagnosa agar berlangsung cepat dan akurat. Dalam periode pertama penelitian ini (2011), telah dikembangkan prototype sistem Computer Aided Diagnosis memanfaatkan teknologi pengo-lahan citra, untuk mengidentifikasi status penyakit Malaria dari citra apusan tipis darah pasien Malaria, sedangkan pada tahun kedua (2012), sistem ini disempurnakan dengan memasukkan modul klasifikasi komponen darah. Sistem yang dikembangkan merupakan transformasi dari pengetahuan yang dimiliki oleh ekspert microscopist ke dalam rule-based classifier dikombinasikan dengan Bayes Decision. Dalam uji coba dengan data citra mikroskopis darah yang dikumpulkan dari berbagai daerah di Indone¬sia, sistem ini menunjukkan hasil yang relatif bagus. Akurasi identifikasi tiap komponen: erythrocyte 94.75%, leucocytes 90%, platelets 41.28% dan erythrocyte yang terinfeksi oleh parasit 84.5%, sedangkan akurasi secara keseluruhan sekitar 77.53%.

Kata Kunci: Computer aided diagnosis, malaria, microphotograph, apusan tipis

I. PENDAHULUAN

penyakit tropis di Indonesia yang harus mendapatkan penanganan serius dan cepat.[1]

Salah satu faktor penting dalam penanganan ka-sus Malaria adalah diagnosa dini Malaria yang akurat. Dalam praktek medis, dikenal berbagai teknik diagnosa seperti PCR, diagnosa klinis dan diagnosa mikroskopi. Di antara teknik tersebut, diagnosa mikroskopi meru-pakan teknik gold standard dan relatif paling murah. Teknik ini memanfaatkan mikroskop untuk menghi-tung parasit, di mana sampel darah diberikan pewar-naan melalui proses kimia yang disebut (Giemsa) stain¬ing. Selain murah, diagnosa mikroskopi juga bersi-fat sensitif, yaitu dapat menentukan kepadatan para

Prosiding InSINas 2012

TI-16 0121: Anto Satriyo Nugroho dkk.

sit dari darah pasien. Diagnosa ini juga informatif, yaitu dapat dipakai untuk menentukan jenis plasmod¬ium Malaria, yang merupakan salah satu pertimban-gan dalam memberikan terapi kepada pasien. Akan tetapi diagnosa mikroskopi ini juga memiliki berba-gai kelemahan. Akurasi diagnosa sangat tergantung kepada keahlian dan pengalaman mikroskopis (orang yang bertugas membaca slide apusan darah). Ter-lebih lagi, prevalensi Malaria di Indonesia cukup ren-dah, sehingga slide yang dibaca umumnya negatif, maksudnya tidak terjangkit Malaria. Padahal untuk memastikan bahwa seorang pasien darahnya negatif jauh lebih sulit daripada menentukan bahwa seorang pasien statusnya positif terjangkit Malaria. Untuk itu, Badan Pengkajian Penerapan Teknologi bekerja sama dengan Lembaga Biologi Molekuler Eijkman dalam pe-ngembangan sistem Computer Aided Diagnosis berba-sis teknik pengolahan citra, yang mampu membantu proses diagnosa, agar berlangsung cepat dan akurat.

Dari kajian literatur, beberapa pihak melakukan upaya serupa untuk kasus Malaria di luar negeri, akan tetapi terbatas pada kasus dengan skala kecil. Misal-nya Premaratne memakai artificial neural network yang dikhususkan pada spesies Plasmodium falciparum. Model tersebut memakai citra sel yang telah dinor-malisasikan sebagai input bagi proses training maupun testing neural network.[2] Studi lain dilakukan oleh Ross yang memakai warna, tekstur dan ciri geometri dari parasit dan sel sebagai vektor fitur, yang selan-jutnya diklasifikasikan ke salah satu spesies plasmod¬ium dalam dua tahap.[3] Akan tetapi studi yang mereka lakukan hanya terbatas pada data berskala kecil, 5 sam-pel untuk training dan 15 sampel untuk testing, sedang-kan karakteristik parasit sangat kompleks. Dapat dis-impulkan bahwa hingga kini, belum ada studi kom-prehensif yang dilakukan untuk mengembangkan com¬puter aided diagnosis terhadap kasus Malaria, khusus-nya di Indonesia berskala besar.

Dalam periode pertama penelitian ini (2011), telah dikembangkan prototype sistem Computer Aided Di¬agnosis memanfaatkan teknologi pengolahan citra, un-tuk mengidentifikasi status penyakit Malaria dari citra apusan tipis darah pasien Malaria. Beberapa algoritma baru pengolahan citra telah dikembangkan oleh tim peneliti dan bekerja sama dengan ahli mikroskopi di Lembaga Biologi Molekuler Eijkman, sehingga hasilnya dapat diverifikasi langsung oleh praktisi lapangan.[4, 5] Pada tahun pertama penelitian, algoritma tersebut telah dievaluasi pada data citrayang telah terinfeksi oleh Plasmodium Falciparum, dengan hasil cukup baik. sen¬sitivity 92.59% dan specificity of 99.65%, Positive Pre¬dictive Value (PPV) of 67.56%, and F1 score of 0.7812.[5]

Dalam tahun kedua penelitian ini (2012), kegiatan penelitian difokuskan pada penyempurnaan prototype system sehingga bisa dipakai untuk mendeteksi keber-

adaan berbagai spesies parasit selain Plasmodium fal-ciparum (P. malariae, P. ovale, P. vivax) , pada citra de-ngan berbagai pada kualitas lapangan.

II. METODOLOGI

A. Persiapan Data Microphotograph

Sample darah pasien Malaria dipersiapkan oleh tim peneliti Lembaga Biologi Molekuler Eijkman, diperoleh dari berbagai daerah di Indonesia. Sample tersebut ter-diri dari 5 spesies: Plasmodium falciparum, Plasmod¬ium vivax, Plasmodium ovale, Plasmodium Malariae, dan Plasmodium Knowlesi.

Setiap slide dilakukan proses staining memakai standard Giemsa staining protocol. Slide kemudian diperiksa di bawah mikriskop dengan pembesaran 10100, dan citra ditangkap dengan 5-megapixel Nikon digital sight DS 5 Mc, yang didesain khusus untuk mikroskop cahaya. Data set tersebut kemudian dianal-isa secara manual, dan dilakukan perhitungan jumlah parasite, serta fase dan tipe masing-masing sel.

B. Sistem Computer Aided Diagnosis

Sistem yang dikembangkan merupakan transformasi pengetahuan yang dimiliki oleh microscopist analyst Lembaga Biologi Molekuler Eijkman ke dalam algo-ritma pengolahan citra. GAMBAR 1 memperlihatkan bagan sistem yang dibangun untuk melakukan iden-tifikasi citra microphotograph darah pasien. Sistem itu terdiri dari fase preprocessing untuk memperbaiki kualitas citra dan menghapus noise, dilanjutkan dengan fase segmentasi komponen darah dan klasifikasi kom-ponen darah. Tujuan dari dua fase ini adalah menge-nali jenis berbagai komponen darah, seperti erythro-cyte, leucocyte, thrombocyte, dan artefacts.

Erythrocyte sendiri dibagi dua: erythrocyte sehat dan suspected erythrocyte, yaitu erythrocyte yang di-curigai terinfeksi oleh parasit. Selanjutnya Bayes Clas-sifier dipakai untuk menentukan apakah erythrocyte tersebut sehat (bercak hitam itu hanya sekedar artefak) ataukah erythrocyte tersebut benar-benar terinfeksi par-asit.

GAMBAR 1: Rancangan sistem Computer Aided Diagno¬sis Malaria

Prosiding InSINas 2012

0121: Anto Satriyo Nugroho dkk. TI-17

B-1. Prepocessing

Dalam tahap ini, citra dikonversikan ke gray-scale image, karena informasi warna tidak diperlukan untuk melakukan klasifikasi komponen sel darah. Noise pada citra dihapus memakai Median filter. Pada GAMBAR 2 diperlihatkan gambar berbagai komponen darah: (1) erythrocyte (sel darah merah), (2) leucocyte (sel darah putih) dan (3) platelet dan (4) erythrocyte yang mengan-dung parasit.

GAMBAR 2: sampel citra darah dengan berbagai kompo-nen yang diberi nomor

B-2. Segmentasi komponen darah

Tahap ini bertujuan untuk melakukan segmentasi, yaitu memisahkan tiap citra komponen darah dari citra latar belakang. Pada citra input, intensitas terbagai ke dalam 3 level: paling terang, menengah dan pa¬ling gelap. Citra thrombocytes, leucocytes dan artefacts umumnya terbentuk dari pixel dengan warna gelap, se-dangkan erythrocytes tersusun dari pixel dengan in-tensitas medium (abu-abu), dan latar belakang paling terang. Karena ada 3 intensitas yang dominan, citra in-put didekomposisikan ke dalam 3 citra memakai Otsu Dual Thresholding. Citra yang diperoleh terdiri dari 3 dan disebut mask. Mask pertama mengidentifikasikan posisi seluruh komponen darah dan dinotasikan de-ngan A.

Mask kedua adalah citra yang terdiri dari pixel

dengan warna paling gelap. Pixel pada mask

ini mengindikasikan posisi artefact, nucleus parasit, thrombocytes dan leucocytes. Mask kedua diidenti-fikasikan dengan Λ.

Mask yang terakhir mengindikasikan posisi erythro-cytes, yaitu yang tersusun dari pixel dengan warna medium. Mask ini dinotasikan dengan E Karena ery-throcytes memiliki struktur cekung, pada saat dual thresholding dilakukan, akan terbentuk lobang di ten-

gah. Agar proses segmentasi berjalan benar, hasil seg-mentasi tersebut diperbaiki dengan proses morfologi hole-filling yang dikembangkan pada penelitian ter-dahulu.[4] GAMBAR 3-4 menunjukkan proses morfologi hole filling yang diterapkan pada sebuah citra, sehingga diperoleh hasil dengan lubang tertutup.

GAMBAR 3: Fast Hole Filling

B-3. Klasifikasi tahap pertama: klasifikasi komponen berbasis aturan

Setelah 3 mask file diperoleh, komponen darah diiden-tifikasi berdasarkan aturan berikut: A mewakili semua komponen dari citra darah,  mewakili komponen dengan pixel gelap, E mewakili komponen dengan intensitas piksel sedang. Lima kelas didefinisikan sebagai berikut: ω1 erythro¬cyte sehat, ω2: leucocyte ω3: thrombocyte, ω4: Thrombocyte menumpang di erythrocyte ω5: suspected erythrocyte. Selan-jutnya klasifikasi dilakukan berdasarkan aturan-aturan seba-gai berikut:[6]

GAMBAR 4: kiri: mask file sebelum dilakukan hole filling kanan: setelah hole filling

(i) X  A s.t. X / X  E X  ω1

(ii) X  A s.t. X  φX  E  φ X  ω3

Prosiding InSINas 2012

TI-18 0121: Anto Satriyo Nugroho dkk.

(iii) X  A s.t. (X  Λ = 0)  (X  E = 0) (τ1  area(X)  τ2)  X  ω4

(iv) X  A s.t. (X  Λ = 0)  (X  E = 0) (area(X) > τ2)  X  ω2

(v) X  A s.t. (X  A = 0)  (X  E = 0) (area(X) < τ1)  X  ω5

B-4. Klasifikasi tahap kedua

Apabila sebuah erythrocyte diklasifikasikan ke dalam ω5 (suspected erythrocyte), maka sel tersebut akan diidenti-fikasikan lebih lanjut untuk memastikan apakah benar-benar terinfeksi oleh parasit atau bercak hitam itu hanya sekedar artefact. Verifikasi ini dilakukan dengan mempartisi sel ke dalam 3 area berdasarkan Otsu dual thresholding (lihat GAM-BAR 5), kemudian menghitung rasio white/black pixel seba-gaimana diperlihatkan pada GAMBAR 6, sebagai acuan untuk membuat keputusan dengan Bayes classifier.[5]

GAMBAR 5: hasil dual thresholding terhadap erythro¬cyte yang terinfeksi oleh parasit dan erythrocyte yang ditumpangi oleh artefacts

GAMBAR 6: rasio white/black pixel sebagaimana se-bagai acuan untuk membuat keputusan dengan Bayes classifier[5]

III. HASIL DAN PEMBAHASAN

Computer Aided Diagnosis system yang dibangun selan-jutnya dievaluasi dengan 40 citra microphotograph. Tingkat

keberhasilan yang dicapai secara keseluruhan adalah 77.63%, dengan akurasi masing-masing: erythrocyte 94.75%, leuco-cytes 90%, platelets 41.28% dan erythrocyte yang terinfeksi oleh parasit 84.5%.[6] Analisa terhadap error menunjukkan bahwa umumnya hal itu disebabkan oleh kegagalan segmen-tasi. Contoh dari kegagalan segmentasi ini diperlihatkan pada GAMBAR 7, di mana thrombocyte masih tersambung dengan erythrocyte, sehingga diidentifikasikan sebagai para-sit. Beberapa kesalahan lain disebabkan oleh komponen yang tumpang tindih (overlapping). Saat ini tengah dikaji solusi dari masalah tersebut dengan memanfaatkan teknik segmen-tasi watershed.[8]

GAMBAR 7: Error yang disebabkan kegagalan segmen-tasi. Thrombocyte masih tersambung dengan erythro¬cyte, sehingga diidentifikasikan sebagai parasit

Eksperimen juga dilakukan terhadap spesies yang lain. Eksperimen ini memakai 5 slide P. Falciparum dan 7 slide P.vivax pada fase asexual. Tiap slide diambil data sebanyak 150 citra microphotograph. Hasil eksperimen ini menun-jukkan bahwa sistem mampu mendeteksi fase aseksual P. fal-ciparum dan P. vivax relatif baik. Untuk P. falciparum, tingkat sensitifitas sistem 73.59% sedangkan specifisitas 98.54%, PPV (Positive Predicted Value) 23.97% dan score F1 sebesar 29.54%. Adapun P. vivax, sensitifitas sistem sebesar 79.15%, spesifisi-tas 98.42%, PPV 7.62%, dan score F1 sebesar 12.69%.[7]

IV. KESIMPULAN

Studi ini merupakan bertujuan untuk mengembangkan sistem Computer Aided Diagnosis, yang membantu praktek klinis di lapangan agar bisa melakukan diagnosa dini Malaria dengan cepat dan akurat. Sistem ini memakai citra micropho¬tograph apusan tipis darah pasien Malaria yang telah dilaku-kan pewarnaan dengan Giemsa staining. Sistem yang diban-gun merupakan transformasi dari pengetahuan ekspert di-agnosa mikroskopis yang sering disebut microscopist pada Lembaga Biologi Molekuler Eijkman, ke dalam algoritma pe-ngolahan citra. Hasil yang diperoleh dibandingkan dengan hasil diagnosa yang dilakukan secara manual oleh para mi¬croscopist tersebut.

Salah satu masalah utama dalam studi ini adalah melakukan klasifikasi komponen darah, dan menentukan ada tidaknya sel darah merah yang terinfeksi. Klasifikasi kom-ponen darah berhasil dilakukan dengan membagi citra darah ke 3 sub citra secara otomatis berdasarkan kemiripan intensi-tasnya, dan selanjutnya mendefinisikan class tiap komponen

Prosiding InSINas 2012

0121: Anto Satriyo Nugroho dkk. TI-19

lewat klasifikasi berbasis aturan (rule based classifier). Secara keseluruhan, algoritma yang dikembangkan mampu mende-teksi 77.53% class tiap komponen darah, dengan akurasi masing-masing: erythrocyte sehat 94.75%, suspicious erythro¬cytes 84.50%. Suspicious erythrocytes ini selanjutnya akan diidentifikasikan dengan Bayes rule, apakah memang benar-benar positif parasit ataukah hanya artefak yang kebetulan posisinya di erythrocyte.

Prototype piranti lunak ini sedang dikembangkan lebih lanjut dengan integrasi pada hardware, yaitu microscope camera dan mikroskop dan disambungkan pada laptop yang ringkas dan dapat dipakai secara portable untuk melakukan identifikasi status Malaria pada daerah rawan. Tahun 2013 di-rencanakan untuk mengevaluasi sistem di kawasan Indonesia timur, yang prevalensi Malaria yang cukup tinggi.

DAFTAR PUSTAKA

[1] US Namru-2 Jakarta (2008), Panduan Pelatihan Diagno¬sis Mikroskopi Malaria, Departemen Parasitologi Medis,

[2] S.P. Premaratne, N.D. Karunaweera, S. Fernando, W.S.R. Pererab, R.P. Asanga, S. Rajapaksha (2003), A Neu¬ral Network Architecture for Automated Recognition of Intracellular Malaria Parasites in Stained Blood Films, APAMI & CJKMI-KOSMI Conference 2003

[3] N.E. Ross, C.J. Pritchard, D.M. Rubin, A.G. Duse, (2006), Automated Image Processing method for the diagnosis and classification of Malaria on thin blood smears, Med¬ical and Biological Engineering and Computing, Vol.44, No.5, pp.427-436, 2006

[4] D. Anggraini, A.S. Nugroho, C. Pratama, I.E. Rozi, A.A. Iskandar, R.N. Hartono (2011), Automatic Status Iden-tification of Microscopic Images Obtained from Malaria Thin Blood Smears, Proc. of 3rd International Conference on Electrical Engineering and Informatics (ICEEI 2011), CDROM A3-2, Institut Teknologi Bandung, Bandung, In-donesia, July 17-19, 2011

[5] D. Anggraini, A.S. Nugroho, C. Pratama, I.E. Rozi, V. Pragesjvara and M. Gunawan (2011), Automated Sta¬tus Identification of Microscopic Images Obtained from Malaria Thin Blood Smears using Bayes Decision: A study case in Plasmodium Falciparum, Proc. of Interna¬tional Conference on Advance Computer Science and In¬formation System 2011, Jakarta, Indonesia, 17-18 Decem¬ber 2011

[6] N.D. Kania, T. Lukito, A.S. Nugroho, I.E. Rozi, M. Gu-nawan, V. Pragesjvara, D. Anggraini (2012), Blood Com-ponent Classification for Malaria Computer Aided Diag¬nosis from Thin Blood Smear Microphotographs, Proc. of 5 International Symposium on Computational Science (ISCS), pp.199-206, 15-16 May 2012

[7] T. Lukito, N.D. Kania, I.E. Rozi, A.S. Nugroho, M. Gunawan, V. Pragesjvara, D. Anggraini (2012), Semi-automated Computer-aided Diagnosis for Malaria Multi species Parasite Detection from Thin Blood Smear Mi¬crophotographs”, Proc. of 5th International Symposium on Computational Science (ISCS), pp.296-302, 15-16 May 2012

[8] A. Riyadi, M.J. Riski, N.N. Faiza, Y. Arti, A.S. Nugroho, M. Gunawan, I.E. Rozi (2012), Evaluasi Algoritma Wa¬tershed pada Pemisahan Citra Apusan Tipis Sel Darah

Prosiding InSINas 2012

Procedural Guide for Ph.D. Students Human Development and Family Science

Auburn University

A PROCEDURAL GUIDE FOR

DOCTOR OF PHILOSOPHY STUDENTS

DEPARTMENT OF HUMAN DEVELOPMENT AND FAMILY SCIENCE

AUBURN UNIVERSITY

updated September 2021

TABLE OF CONTENTS

Introduction

Appointing a Temporary Advisor

Selecting a Major Professor

Forming an Advisory Committee

HDFS Graduate Faculty

Plan of Study

Graduate School Course Requirements for the Plan of Study

Professional Socialization Guidelines

Lines of Authority

Annual Evaluation

Steady Progress toward the Degree

Evaluation Process

Evaluation Outcomes

HDFS Assistantship Policy

M.S. Ph.D. Students

Ph.D. Students

Graduate Assistant Evaluation Policy

Doctoral Competencies

The Doctoral Qualifying (General/Preliminary) Examination

Guidelines for the Qualifying (General/Preliminary) Exam

The General Oral Exam (Defense of the Written Exam)

Dissertation and Proposal Guide

Timing of the Dissertation Proposal Meeting

Guidelines for Working with Your Committee

Requirements for a Proposal

Elements of a Dissertation

Final Examination (Dissertation Defense)

Revised 10/2018

Procedural Guide for Ph.D. Students Human Development and Family Studies

Auburn University

INTRODUCTION

The Doctor of Philosophy is conferred in recognition of the mastery of a special field of learning as shown by the satisfactory completion of a prescribed course of study and investigation, the successful passing of general examinations covering the major and minor fields, the preparation of an acceptable dissertation reflecting high achievement in scholarship and independent investigation, and the passing of a final examination on the dissertation and related subjects. The degree is a research degree not conferred merely upon fulfillment of technical requirements, but awarded in recognition of the ability to think and work independently, originally, and creatively in a chosen field.

(Italicized material throughout the text of the guide has been selectively drawn (or edited) from the graduate school web page and indicates stated policies or definitions of the Auburn University Graduate School. For more complete information go to http://www.grad.auburn.edu/.)

During your first year in the Ph.D. program in the Department of Human Development and Family Science (HDFS) you will make a number of important decisions. These decisions may influence not only the direction of your doctoral training but also the course of your early career. First, you will select your major professor and the advisory committee that will direct your studies and examine your progress at several critical points during your program. Second, you will develop your tentative plan of study to identify the courses you will take as electives and supporting emphases. Third, you will begin planning and pursuing the specific skills and experiences needed to prepare for your career. This guide is intended to clarify the purposes of, and procedures associated with, each of these decisions. Ultimately, you are responsible for learning and following HDFS and Graduate School policies and procedures.

APPOINTING A TEMPORARY ADVISOR

If you have entered the doctoral program in HDFS without coming through our master’s program and you have not specified a faculty member as your preferred major professor, the graduate program officer (GPO) will serve as your temporary advisor. As temporary advisor, the GPO will advise you regarding your course of study in the first semester. During the first semester, we encourage you to identify a faculty member to serve as your major professor. The process of selecting a major professor and advisory committee is important and addressed in the following sections.

SELECTING A MAJOR PROFESSOR

You may select a major professor at any time during your first semester of study; it is important to select your major professor no later than the second semester of your first year. Your major professor must be a member of the Graduate Faculty and eligible to direct doctoral students. In making your selection, you should set individual appointments with at least two faculty members who appear to have professional and research interests congruent with your own. During the appointments, you should evaluate the fit between your own interests and the relevant professional and research interests of each potential major professor. In addition, you should discuss with each professor their style of mentoring. Beyond questions you arrive at on your own or in conversations with other students who have already made these decisions, consider seeking answers to the following questions:

* How many students is the faculty member working with? (Sometimes having several students means an advisor will have limited time; sometimes it means the advisor is especially effective as a mentor. Get the names of some of the professor’s advisees and talk with them before committing to an advisor.)

Revised 10/2018

Procedural Guide for Ph.D. Students Human Development and Family Studies

Auburn University

* How connected are students to the professor’s research program? (Put differently, will you work with the professor on his/her own research program or will you be allowed [required] to pursue an interest that is entirely your own? There are advantages and disadvantages of each approach. Talk about them. Talk with current advisees of the professor you are considering.)

* How available is the professor to students?

* What kinds of opportunities are available to students working with the professor? (To the extent that you know what you want to do with your degree, you will also be able to identify [with the assistance and advice of your major professor and advisory committee] experiences that will promote your capacity to be competitive for that kind of career. Will you have opportunities to get those experiences while working with the professors you are considering? Talk about the kinds of teaching experiences you will get while working with the professor. What kinds of research experiences will you get? Will there be publication opportunities? Grant writing opportunities? What kinds of opportunities will you get to apply your knowledge in the service of others or in a policy context?)

We encourage you to choose your major professor carefully. Before making the decision, consider all appropriate faculty prospects and talk with other students who have worked with the professors you are considering. You can identify possible major professors by reviewing information about faculty research interests on the HDFS web page. In addition, the introductory seminar and department events should facilitate your awareness of well-matched professors. Finally, a discussion with the department head or GPO can be helpful in directing you to faculty whose interests are aligned with yours.

Doctoral students sometimes wish to change their major professor, or major professors may want to discontinue their role with a student. On such occasions, the student and advisor are encouraged to meet to discuss their wishes and attempt to arrive at a mutually acceptable arrangement. The GPO and department head are available to help if concerns arise. An important point is that students cannot progress in the program without a major professor. Hence, discontinuation from one major professor must be accompanied by realignment with another major professor for the student to continue in the program.

FORMING AN ADVISORY COMMITTEE

You should form an advisory committee after selecting a major professor. The advisory committee must be

selected in consultation with your major professor. We encourage you to select the advisory committee during the

first year of the doctoral program.

When selecting an advisory committee, consider the theoretical/methodological fit between your major professor and the prospective committee members, the general interest domains and specific content areas of the

prospects, and the degree to which you and the major professor work effectively with the prospective members. Your major professor’s input is critical in the selection of the advisory committee.

The Graduate School indicates that: The advisory committee is responsible for developing the student's plan of study and conducting the doctoral general and final examinations. It should consist of at least four members of the graduate faculty. A majority of the Auburn University affiliated committee members, including the major professor, must be members of the Graduate Faculty, Level 2. The major professor must also be a graduate faculty member of the program granting the degree. Faculty eligible for directing doctoral students as major professor have considerable experience advising thesis or dissertation research, they have considerable research experience themselves, and they remain active in the process of research. Although the majority of advisory committee members must share this highest standing on the Graduate Faculty (Level 2), the primary criterion for the selection

Revised 10/2018

Procedural Guide for Ph.D. Students Human Development and Family Studies

Auburn University

of advisory committee members is the fit between the student’s dissertation interest and the faculty member’s expertise. The goal is to ensure that each doctoral student has the best combination of advisors and examiners available given his/her research interests. The student and major professor should collaborate in identifying the committee that fulfills this goal.

HDFS Graduate Faculty:

Francesca Adler-Baeder, Ph.D.

Katrina Akande, Ph.D.

Cory Cobb, Ph.D.

Adrienne Duke, Ph.D.

Mona El-Sheikh, Ph.D.

Stephen Erath, Ph.D.

Cynthia Frosch, Ph.D.

Thomas Fuller-Rowell, Ph.D.

Ben Hinnant, Ph.D.

Jennifer Kerpelman, Ph.D.

Scott Ketring, Ph.D.

Kyle Kostelecky, Ph.D.

Mallory Lucier-Greer, Ph.D.

Julianne McGill, Ph.D.

Joshua Novak, Ph.D.

Lauren Ruhlmann, Ph.D.

Diana Samek, Ph.D.

Wendy Troop-Gordon, Ph.D.

Brian Vaughn, Ph.D.

Silvia Vilches, Ph.D.

Angela Wiley, Ph.D.

Elif Dede Yildirim, Ph.D.

Some students may wish to have members on their advisory committee from outside Auburn University, such as a scholar from another university. Graduate school policy indicates that: Affiliate faculty and other scholars not employed by the University may serve on individual advisory committees with the approval of the Dean of the Graduate School. Such approval does not imply Graduate Faculty membership. The department head or other appropriate administrative officer should forward a recommendation for approval along with the scholarly credentials of the individual to the Graduate School for review and subsequent action.

The formal appointment of the advisory committee occurs when the Committee, Transfers, Exceptions, and Candidacy (CTEC) form (http://graduate.auburn.edu/current-students/committee-selection/) is approved by the Graduate School. Any change in the membership of a student’s advisory committee must be reported to the Graduate School via email with copies to the former and new committee members as well as the committee chair.

PLAN OF STUDY

Following selection of the major professor and advisory committee, the next step is to submit the CTEC form (CTEC; http://graduate.auburn.edu/current-students/committee-selection/). The purpose of the form is to identify the student, committee chair/major professor, committee members, and any transfer courses or curriculum

Revised 10/2018

Procedural Guide for Ph.D. Students Human Development and Family Studies

Auburn University

exceptions or substitutions. Transfer courses and curriculum exceptions or substitutions should be discussed and approved in advance by the committee chair/major professor. Doctoral students must submit and obtain approval of the CTEC form prior to submitting the Request for the Report on the General Oral Examination (i.e., before scheduling the qualifying exam defense with the Graduate School).

Graduate School Course Requirements for the Plan of Study

The minimum number of hours in a doctoral program is 60 semester hours earned through instruction beyond the bachelor’s degree, including 1) a minimum of 30 semester hours of graded (e.g., A, B) graduate course work (6000-level and above) and 2) a minimum of 30 semester hours of additional graduate course work (6000-level and above) that may include ungraded courses, 7990 (Thesis Research) and 8990 (Dissertation Research) and must include at least 10 hours of 8990. Some programs require more than 60 semester hours, and requirements may vary according to a student’s background and interests.

The total number of credit hours that may be transferred from another accredited institution toward a doctoral degree varies by program but must be less than 50% of the credit hours listed on the Plan of Study. Such transfer credit 1) must fall within the time limits of the degree, and 2) must be approved by the advisory committee and the Dean of the Graduate School. A maximum of 4 hours of 7990 from a completed master’s program may be counted. You may enroll in 8990 at any time during your program that your advisor considers appropriate. However, the number of 8990 hours taken in any given semester should reflect the amount of time being spent that semester on the dissertation and the degree to which university resources are being utilized. No grade is assigned to 8990 hours.

PROFESSIONAL SOCIALIZATION GUIDELINES

Working toward the Ph.D. is qualitatively different from working on an undergraduate or even a master’s degree. You are expected to work more independently (i.e., in a more self-motivated, self-directed way) and to perform at a higher level, producing better-quality work. At this stage in your scholarly development you are being trained as a professional colleague. Thus, over your doctoral program you should increasingly identify yourself as a professional rather than a student. Success in the Ph.D. program in HDFS, and later in the larger professional world, requires more than just completing a specified number of courses, passing exams, and completing a dissertation. It involves learning to work with other professionals, developing leadership skills, and gaining knowledge of professional protocols, as well as building research and academic skills. To promote these competencies, your committee will recommend experiences and provide feedback regarding research, teaching, and service. To develop professional relationship skills and to acquire knowledge of basic professional protocol, you will have opportunities to observe and experience various professional situations, again at the recommendation of your advisor and committee.

Good time management is essential to accomplishing all that will be required of you. You will need considerable time to focus on your coursework, but time to interact with fellow classmates and faculty outside the classroom is also important. Much of your learning will be informal and spontaneous and some of it will be vicarious as you observe the efforts and outcomes of other students in the program. Doctoral students are strongly encouraged to attend all HDFS public defenses of qualifying exams and dissertations when possible. Observing other students is valuable, as it familiarizes you with the process and provides performance models with both strengths and limitations. Although much of the feedback you will receive from faculty and fellow students regarding your development will be informal and spontaneous, the department also has a formal evaluation process that will provide direct feedback to you on an annual basis. The goal of both formal and informal feedback is improvement

Revised 10/2018

Procedural Guide for Ph.D. Students Human Development and Family Studies

Auburn University

and professional socialization.

The HDFS faculty want graduates of our program to be strong and competitive new professionals on the job market. Our evaluations of your performance in this department, in classes and assistantships but also in the development of your scholarly specialization and in other aspects of your professional socialization, will be reflected in letters of reference, recommendations for awards, etc. We want you to be successful.

LINES OF AUTHORITY

There may be times when you have a concern regarding a course, some aspect of the doctoral program, doctoral requirements, your assistantship, or some other issue. For concerns regarding a course or assistantship, we encourage you to first discuss the concern with the faculty member who is teaching the course or supervising the assistantship. If the concern is not satisfactorily addressed, then you should bring the concern to the GPO or Department Head. If the concern does not deal with a course or assistantship and the student does not know where to begin in order to address the concern, contact the GPO or Department Head.

ANNUAL EVALUATION

As a doctoral student, your progress in the doctoral program will be evaluated on an annual basis.

Steady Progress toward the Degree

There is so much to do as you prepare for your career that it is easy to lose track of time. Students must understand that their success in the program is defined fundamentally in terms of the progress they are making toward the degree. Progress involves completion of the core coursework, electives, the MS thesis (if relevant to the student), the qualifying exam, and the dissertation, as well as the skill-building experiences recommended for students by the advisor and committee.

The Ph.D. is conceptualized as a 5-or-6-year degree program. This timeframe assumes that the student will take the full doctoral core of coursework including that which is in common with the MS degree. Therefore, the timeframe applies not only to students who come to the program with a bachelor’s degree and intend to proceed all the way to the Ph.D., but also to those students who complete an MS degree elsewhere and come to Auburn to pursue the Ph.D. All doctoral graduates will complete the full core unless they can show a transcript and syllabus that indicates the content of a graduate class taken elsewhere covers the content of a given class. The teaching faculty member decides whether the class content has been adequately addressed.

The 5-or-6 year timeframe is important as a standard against which to measure a student’s progress toward the degree. In order to accomplish the degree within this timeframe, students must keep their sights on completing the intermediate steps required by the degree. If adequate progress is not made, the assistantship support available to the student will be curtailed. Departmental assistantships and university fellowships are not available to students who are not making adequate progress toward the degree. Inadequate progress may or may not result in expulsion from the program, but if not, financial support will be the responsibility of the student not advancing at pace. Students for whom assistantship support is important to the ability to pursue graduate education must ensure that they progress appropriately.

Revised 10/2018

Procedural Guide for Ph.D. Students Human Development and Family Studies

Auburn University

Evaluation Process

Early in the spring semester, the GPO will direct you to complete a form outlining the progress you have made toward your degree goals. As part of this process, you will also declare your goals for the coming year in terms of research, teaching, and outreach/service. There will also be an opportunity for you to identity longer-range goals and to engage in self-evaluation in the context of your program. Along with this form, you will submit an updated curriculum vita. Other supporting documents such as teaching evaluations, extension/outreach products, conference presentations, and published papers that were generated during the evaluation year should be submitted with the vita. Evaluations of your assistantship performance (submitted by your assistantship supervisor) are considered in the review as are comments from your instructors and faculty who have had other exposure to you as a student in the program.

The goal of the evaluation process is to assess each student’s progress and accomplishments and to identify concerns or areas for improvement so that supportive action can occur. For students who are on track, suggestions about suitable opportunities may also be raised, and student accomplishments are brought to the attention of the faculty. Students have been nominated for significant student awards as a result of the annual review.

Evaluation Outcomes

Students will receive one of three overall ratings from the review.

• “Satisfactory progress” indicates the student is progressing as expected or better.

• “Concerns with progress” means the faculty have identified specific issues that must be addressed in the coming year. Students who receive a rating of "concerns" are funded as TAs or RAs only if their major professor is willing to provide funding for them or vouch for their progress the next year. Students for whom the faculty have concerns in consecutive years receive the rating “unsatisfactory” in the second year.

• “Unsatisfactory progress” indicates the faculty believe previously identified concerns have not been adequately addressed and/or the student is seriously off track in the program. Students with this rating will not be eligible for assistantship support for the coming year. A student rated "unsatisfactory" in one year must attain a rating of "satisfactory" during the following year to continue in the program.

Following the faculty meeting, your major professor will compile faculty feedback in a letter and share it with you. The letter becomes a part of your permanent record in the program.

HDFS ASSISTANTSHIP POLICY

All assistantship funding in the Department of Human Development and Family Science is contingent upon remaining in good standing in the program, which means that the student is making satisfactory progress toward the degree. Furthermore, all funding is dependent upon departmental need, the availability of funds, and the approval of the fiscal year budget by the Auburn University Board of Trustees.

Assistantship stipends are paid on different scales for M.S. versus Ph.D. students.

M.S. - Ph.D. Students

For students who originally entered the M.S. program and transitioned to the Ph.D. program, the pay scale remains at the M.S. level until the semester following the successful defense of the master’s thesis.

Revised 10/2018

Procedural Guide for Ph.D. Students Human Development and Family Studies

Auburn University

Because third-year M.S. students tracking into the doctoral program are taking doctoral core courses while completing their thesis, they are not considered off track in the doctoral program as long as they complete the thesis by the end of their third year in the program. A maximum of six years of departmental support (including any provided while an M.S. student in HDFS) is available to students whose annual reviews indicate that they have remained on track throughout their program.

Ph.D. Students

Students who joined the doctoral program with an M.S. degree from another program receive a stipend at the doctoral level of pay. They are required to take the full doctoral core unless they are able to transfer core classes from their master’s program but may transfer electives and may not need to complete a thesis (e.g., if they have already completed a thesis); thus, a maximum of five years of departmental support is available to students whose annual reviews indicate they have remained on track.

GRADUATE ASSISTANT EVALUATION POLICY

Each semester you hold a graduate assistantship, you should expect to meet with the supervisor of the assistantship at the beginning of the semester to discuss your duties and the advisor’s expectations. You should also expect at least one meeting, typically at the end of the semester, during which your supervisor reviews your assistantship performance. Typical items in the evaluation include whether you were: timely in your performance of assistantship duties, thorough and thoughtful in your role, responsive to feedback, able to take the lead in appropriate times and ways, a team player in relation to the people you work with in your duties, and developing appropriate competencies related to the assistantship.

There is a separate evaluation form for each type of assistantship. You should expect to see your supervisor’s evaluations and have the opportunity to discuss them in a meeting with her/him. These forms become part of your permanent record in the program and are also part of your annual evaluation. If you have concerns about your performance in an assistantship, you should approach your supervisor for feedback when your concerns arise rather than waiting to the end of the semester.

If you are the instructor of record for a class, the teaching evaluations completed by your students will also be part of your evaluation. They should be included with the material you provide for the annual review.

DOCTORAL COMPETENCIES

Doctoral students join the HDFS program with different goals, so the competencies or skills that they hope to develop in the program are likely to vary to some degree. However, the HDFS program has some goals in common for all doctoral students. We want all doctoral graduates to have the ability to perform competently as teachers, researchers, and providers of professional service or outreach. Thus, in addition to performing well in class, you will be asked to address the competencies below in your annual evaluation as you plan and implement your program and professional goals.

Teaching Competencies

1. Create appropriate materials for an HDFS course including course syllabi, homework, and tests.

2. Plan lectures relevant to an HDFS course and/or appropriate material for an entire HDFS course.

Revised 10/2018

Procedural Guide for Ph.D. Students Human Development and Family Studies

Auburn University

3. Implement academic honesty policies and ethical treatment of students.

4. Teach effectively.

Research Competencies

1. Conceptualize an HDFS-relevant research question and develop an appropriate research design.

2. Collect data.

3. Apply statistical software packages to independently to analyze data.

4. Develop an analysis strategy and interpret research data independently.

5. Prepare and submit manuscripts for publication.

Service Competencies

1. Participate in university, college, department, or community outreach.

2. Provide professional service including membership in professional organizations.

THE DOCTORAL QUALIFYING (GENERAL/PRELIMINARY) EXAMINATION

Before doctoral students can be admitted to candidacy for the Ph.D. degree, they must pass the written and oral phases of the qualifying (or general or preliminary) examination. The qualifying exam process may begin early in the student’s program but the qualifying exam cannot be defended until the student has completed all required (core) coursework in the doctoral program. Each doctoral student in HDFS takes a core of coursework intended to provide breadth in the student’s understanding of our multi-disciplinary field and to integrate its many threads. In addition, a prominent aspect of doctoral education is the development of depth in a specialization area that will serve as the foundation of the student’s professional career and contribution to the field. Both this breadth and depth occur as students pursue their coursework and their varied assistantship, practicum, and independent learning experiences working with other students and professors in the department. The qualifying (general/preliminary) examination in HDFS represents an opportunity for students to demonstrate the breadth and depth of their knowledge in an independent fashion as they take the lead in producing one of three research products (described below) in collaboration with their major professor or another professor (with the approval of the major professor). This product then becomes the primary focus of the oral component of the qualifying exam. Only after both written and oral parts of the qualifying (general/preliminary) exam have been passed is a student admitted to candidacy for the Ph.D. degree. Students must be admitted to candidacy before their dissertation proposal meeting.

The following is edited from the Graduate School’s policy regarding the general oral exam:

A general examination, often called the “preliminary examination,” is required of all applicants for the degree of doctor of philosophy. It consists of written and oral testing by the student’s advisory committee (or by an examination committee designated by the student’s academic program) in the student’s major and minor. The written portion of the examination does not require approval in advance by the Graduate School. The oral portion, however, does require such approval. Arrangements for the oral examination must be made by application to the Graduate School at least one week in advance of the examination. The primary purpose of the general examination is to assess the student’s understanding of the broad body of knowledge in a field of study. The examination also affords the advisory committee an opportunity to review the student’s proposed research and understanding of research methods and literature in the chosen field. If the general examination reveals deficiencies in any of these areas, the advisory committee may recommend remedial work, re-examination, or discontinuation of doctoral

Revised 10/2018

Procedural Guide for Ph.D. Students Human Development and Family Studies

Auburn University

study.

The general oral examination should be conducted immediately after the successful completion of the written examination and well before the final examination. At least one complete semester must intervene between the general oral and final examinations. The two examinations thus cannot be taken either in the same semester or in consecutive semesters. Successful completion of the oral examination requires unanimous support of the student’s advisory committee. If the general oral examination is failed, a re-examination may be given on recommendation of the committee and approval by the Dean of the Graduate School. Further examinations require exceptional circumstances and approval by the Graduate Council.

The student becomes a candidate for the degree on successful completion of the general examination. Guidelines for the Qualifying (General/Preliminary) Exam

The qualifying exam serves as an assessment of the student’s developing research competencies and involves generating one of the following scholarly products: a critical review paper of publication quality, an empirical paper of publication quality (that is not the master’s thesis), or a grant proposal of sufficient quality to submit for funding. Students are encouraged to submit the scholarly product for publication in a peer-reviewed journal or to a funding agency; however, submission is not required as part of the qualifying exam process. The target journal or funding mechanism should be appropriate for a new professional in the field and should be specified for the committee by the student when the written proposal is submitted. More than one possible journal or funding mechanism may be specified.

The student’s choice of research topic and research product should be considered in close consultation with the major professor and should represent the student’s area of specialization. The scholarly product should also be consistent with the student’s career goals. For example, the product may address basic or applied research topics and may present implications for future research, practice, or policy. Grant proposals may be written for submission to a public or private research agency or to a service foundation.

After choosing a research topic and product in close consultation with the major professor, the student completes the written proposal, written exam, and oral exam, each described below.

Written proposal

The written proposal serves as a research proposal as well as an initial assessment of the student’s research competencies. It allows the advisory committee to identify strengths and areas that may require further development in the student’s doctoral training and to provide recommendations about the written exam.

The written proposal challenges the student to think and write independently, though reflective guidance from the major professor is encouraged. In other words, at this stage of the qualifying exam process, the student generates ideas (e.g., about theoretical frameworks, hypotheses, measures, planned analyses), and the major professor helps the student understand the strengths and limitations of the ideas. The major professor also provides feedback about the content and style of writing, but the student writes the entire proposal.

The written proposal introduces the topic and its significance, identifies the key theories and studies, and outlines the innovations and questions or hypotheses of the proposal or paper. This introductory section should span approximately the same number of pages as the Introduction section of a standard journal publication (i.e., 5-10

Revised 10/2018

Procedural Guide for Ph.D. Students Human Development and Family Studies

Auburn University

pages). In addition, unless the student proposes a critical review paper, the written proposal includes a standard Method or Approach section with appropriate sub-sections (e.g., participants, procedures, measures, and planned analyses).

Upon completion of the written proposal, the student arranges a meeting with the advisory committee and submits the written proposal to the advisory committee at least one week before the meeting. The student begins the meeting with a brief presentation of the proposal (e.g., 10-15 minutes), followed by a period during which the advisory committee asks questions and provides recommendations for the written exam. Within one week following the proposal meeting, the student should write a brief summary of the discussion with an emphasis on decisions or modifications to the original proposal and email the summary to the committee for their records and reference when reviewing the full written exam.

Written exam

Preparing the written exam is an important learning and professional development experience. The written exam serves as a research product as well as a reflection of the student’s research competencies. Along with the oral exam, the written exam helps the advisory committee determine the student’s readiness for advancement to doctoral candidacy and the dissertation.

As compared to the written proposal, the major professor may play a more active role in the development of the written exam (e.g., flesh out the approved study plan, assist with troubleshooting analyses, and assist with processing discussion points), though it remains essential that the student produces the written exam in the lead role. Providing feedback through verbal or written comments is preferred over text editing and writing. Given the opportunity for increased input from the major professor, the benefit of feedback from the advisory committee and other colleagues, and additional time for independent learning and skill-building, the written exam should demonstrate progress in the student’s thinking and writing beyond the written proposal.

In addition to the scholarly product of choice, the written exam may be accompanied by a memo explaining any revisions to the full written product that deviate significantly from the approved proposal.

Upon completion of the written exam, the student submits the written exam to the advisory committee and tentatively schedules the oral exam with the advisory committee no less than three weeks from the submission of the written exam. The advisory committee reviews the written exam within one week. Possible outcomes of the committee review include: accept without revision or with minor revisions (which requires unanimous agreement) or major revisions required. The student is not permitted to proceed to the oral exam unless every member of the advisory committee accepts the written exam without revision or with minor revisions. Major revisions must be specified clearly by committee members through feedback about limitations as well as recommendations about how the exam could be improved. The student should prepare a revision for the committee following the assessment by even one advisory committee member that major revisions are required. Up to two rounds of revision and committee review are possible. Each revision should be accompanied by a detailed response memo that lists each committee member's questions and suggestions and an explanation about how they were addressed. If two rounds of revisions do not yield a product that is considered acceptable without revision or with minor revisions by all committee members, the written exam is considered failed and the student’s graduate program is concluded.

Revised 10/2018

Procedural Guide for Ph.D. Students Human Development and Family Studies

Auburn University

The General Oral Exam (Qualifying Exam Defense)

The oral defense of the written exam occurs within one month of receiving an acceptable outcome on the written exam. The Graduate School requires the student to apply for the oral exam (Request for the Report on the General Oral Examination) at least two weeks in advance of the oral exam. Thus, the oral exam should be scheduled with the advisory committee no less than three weeks from the submission of the written exam to the advisory committee, allowing sufficient time for the student to submit the application for the oral exam to the Graduate School (if the written exam is accepted by the advisory committee) or to reschedule the oral exam with the advisory committee (if the advisory committee requires revisions to the written exam before proceeding to the oral exam). The time and place of the oral exam, as well as the members of the advisory/examination committee, are identified on the application form. The application is available at the following website: http://graduate.auburn.edu/current-students/forms-directory/.

When the time and location of the defense is finalized (i.e., scheduled through the Graduate School), the major professor should announce the defense to HDFS faculty and graduate students. Students preparing for the oral exam are strongly encouraged to consult with their major professor, their committee members, or any other relevant resource.

The general oral exam has two phases. First is an approximately 45-minute, professional-level presentation of the written exam to the HDFS faculty and graduate student body. This presentation is followed by a question-and-answer period during which the student fields questions from the audience about the presentation. Approximately one hour should be allocated for the presentation and Q-A session. The audience is then dismissed and the second part of the exam proceeds. (Note, however, that Graduate School policy authorizes any graduate faculty member to attend the second part of the defense.) Although the second part of the oral exam typically continues to address the written exam, any aspect of the student’s specialization or any area of the Ph.D. core or program is legitimate examination material. The purpose of the oral examination is not simply to assess the student’s knowledge base, but also to test the student’s capacity to think extemporaneously.

At the end of the meeting, the student is typically asked to leave the room for a few minutes while the committee discusses the student’s performance and considers their votes. The votes are ‘pass’ or ‘fail’. To pass the exam, the committee vote must support passing unanimously. Committee members register their vote through the Graduate School’s online form, which is automatically delivered to committee members around the time of the exam. Although there is no official notice, students who have passed the general oral examination can change their designation from ‘doctoral student’ to ‘doctoral candidate’.

In the event of a failure of the general oral exam, students are allowed one retake. Before a failed oral exam ends, the committee’s expectations for the student’s performance in a retake and their recommendation for the timeframe should be clearly specified. One (and only one) retake of the oral exam is permitted, and the student should have a minimum of four weeks to prepare for it. When the time and place for the retake are defined, the student must reapply to the Graduate School for the exam. Only the student and committee reconvene for the retake. The committee should specify at the end of the first oral exam whether a presentation will be part of the second oral defense.

A second failure of the oral exam results in dismissal from the program.

Revised 10/2018

Procedural Guide for Ph.D. Students Human Development and Family Studies

Auburn University

DISSERTATION AND PROPOSAL GUIDE

Timing of the Dissertation Proposal Meeting

You must pass the General Oral Examination before the dissertation proposal meeting.

Guidelines for Working with Your Committee

The HDFS Graduate Advisory Committee has developed the following guidelines to help you in the preparation of your proposal and dissertation.

The purpose of a dissertation is to help you further develop skills in all aspects of the research process, which may include conceptualizing a research question, developing a research design, collecting data, coding and analyzing data, and interpreting results. The dissertation provides a structure through which these skills may be improved. Your committee, therefore, will expect you to demonstrate skill in each of these areas. It is the prerogative of the committee to ask questions regarding any or all of these skills.

You should seek advice from each member of your advisory committee when preparing and writing your dissertation proposal. Each committee member should receive a draft of your proposal at least one week before the meeting with your doctoral advisory committee at which you discuss and defend your proposal. Different committees may need more time, and it is their prerogative to request it.

As you proceed through the process of completing the dissertation, it is your responsibility to check in the Graduate School Bulletin for the graduate school deadlines (http://graduate.auburn.edu/calendar/). The format-check deadline is early in the semester. The deadline for submitting the final dissertation is usually around mid-semester for students intending to defend their dissertation and graduate in the same semester.

Requirements for a Proposal

A dissertation proposal is the plan of action that results in a final dissertation. As such, the Introduction and the Review of Literature will remain largely the same from proposal to dissertation. In most cases, the Method section will be changed only by conversion from future to past tense. Thus, the proposal is not a draft that will be polished up later. Students should understand, however, that many dissertations will confront unanticipated problems that will necessitate substantive changes to the proposal before the final draft of the dissertation is submitted.

You should consider the proposal defensible (i.e., ready for an examination meeting with the advisory committee) only when the Introduction, Literature Review, Method, References, and appropriate Appendices are finalized in the judgment of your major professor.

All research with human subjects must be approved in advance by the Institutional Review Board for the Use of Human Subjects in Research (IRB). The department head must approve protocols before they are forwarded to the IRB.

Elements of a Dissertation

The APA Publication Manual provides guidance for all of the sections that will appear in a dissertation with the exception of the Review of the Literature and Proposed Analysis sections (or see http://www.apastyle.org for the

Revised 10/2018

Procedural Guide for Ph.D. Students Human Development and Family Studies

Auburn University

same information). In the following paragraphs, we provide a brief description of elements of a typical dissertation and note some of the ways in which the AU Thesis guidelines differ from APA practice. Please use the APA Manual (http://www.apastyle.org) and the AU Guide to Preparation and Submission of Theses and Dissertations (http://graduate.auburn.edu/current-students/electronic-thesis-dissertation-guide/) throughout preparation of your document.

A dissertation normally consists of:

Prefatory Material: This material given in the opening pages of the dissertation is stipulated by the Auburn University Thesis and Dissertation Guide (certificate of approval, title page, table of contents, etc.)

Introduction: The Introduction is a brief statement of the question and an overview of the logic for the development of the question (including supporting literature). It may help to think of the Introduction as a section that could (with little editing) become the introduction to a published article. Specific hypotheses may be presented in the Introduction or may be held for the Literature Review section (a decision you should reach with your major professor).

Review of the Literature: The Review of the Literature is an extensive review of extant literature on all topics directly related to your dissertation question (e.g., theoretical and empirical background, a discussion of conceptual issues related to measures, any critique of previous literature and/or methodologies). The Literature Review should build logically to a statement of the specific research hypotheses or questions that will guide the analytic strategy for your study.

Method: The Method section is organized by subheadings that provide information about the following topics: Participants, Procedure, Measures, and Proposed Analyses. A thorough description of participants in your study is presented in the Participants section. This description includes all attributes of the sample that will be relevant to the analysis (e.g., gender and racial composition, age, socioeconomic status of subjects) but also relevant attributes of subjects that may not be included as variables in the analysis (e.g., region of country where subjects reside, number of subjects contacted to participate, number refusing). The Procedure details the steps undertaken to collect the data that you will analyze. The purpose of the Procedure is to permit a critical review of your data collection strategy and to provide enough information for others to replicate your study. The Measures section includes information about how each variable is measured. Estimates of reliability and validity for all measures are important pieces of information and should be included for each variable where relevant. Where possible, these estimates should be based in the data collected for your study. Furthermore, since many studies use previously existing measures for some or all variables in the analysis, published reports of the reliability and validity of these measures based in previous research is also legitimate and important information. The Proposed Analysis specifies each step in the analysis that will be used to answer the questions or hypotheses of the study. The presentation of these analyses, therefore, should link directly to the research questions or hypotheses.

Results: At a minimum, the Results section is a presentation of the outcome of all the proposed analyses. It may be organized (a) by research questions or hypotheses, (b) by issue or content area, or (c) some other logical criterion. Sometimes the planned analyses raise other interesting questions, and thus, the Results section may contain analyses beyond those originally planned.

Discussion: The Discussion is a brief presentation of each finding or group of findings together with explanations and alternative explanations of their meaning. The Discussion should be related to the

Revised 10/2018

Procedural Guide for Ph.D. Students Human Development and Family Studies

Auburn University

Introduction by connecting the findings with the driving issues that originally sparked the project. Furthermore, implications, limitations, and caveats of the results should be considered.

References: Students should use APA guidelines for listing references.

Footnotes: According to the Graduate School, footnotes may be placed at the bottom of the relevant page or the end of the section in which the footnote appears.

Tables and Figures: Tables and figures are integrated into the text near their first citation for AU dissertations and theses. See AU Dissertation Guide and APA for guidance on Table and Figure construction.

Appendices: Any pertinent material (e.g., informed consent letter, copies of questionnaires or surveys, interview protocol, observational coding forms, additional analyses not presented in Results) should be contained within separate appendices identified alphabetically at the end of the thesis/dissertation.

Multi-Study Dissertation Option. A multi-study (multi-publication) dissertation is an alternative to the single-study dissertation. In addition to the prefatory material described above and standard-length sections of a scientific article (Abstract, Introduction, Method, Results, Discussion, References, Tables and Figures), the multi-study dissertation requires General Introduction and General Discussion sections that situate the dissertation studies within an overarching topic area and explain how the studies separately and jointly advance knowledge on the overarching topic. The General Introduction and General Discussion sections span approximately the same number of pages as standard Introduction or Discussion sections of scientific articles (i.e., 5-10 pages). Typical journal space guidelines limit the extensiveness (breadth) of the literature reviews in the multi-study dissertation, but this option raises the challenge of integrating the dissertation studies. A paper that is published between the dissertation proposal and the dissertation defense is permissible as part of the multi-study dissertation as long as the paper is student-led and the research plan for the published paper was proposed and approved as part of the dissertation proposal process.

The dissertation may be accompanied by a memo explaining any revisions that deviate significantly from the proposal.

Final Examination (Dissertation Defense)

After a draft of your dissertation has been completed and approved by your advisory committee, it is submitted to the Graduate School. The representative of the Graduate School (the university reader) then reviews the dissertation. This review process adds weeks to the timetable for dissertation completion and should be accounted for as students plan their final semester. (It is common, but not required, for the university reader to be recruited by the student and/or major professor well in advance of the completion of the dissertation. Identifying the university reader in advance can reduce the time between your committee’s acknowledgment of the readiness of your dissertation to be defended and your actual defense.) When the university reader has approved the dissertation, you should apply for the final examination at least one week prior to the planned meeting.

Like the oral defense of your general written exam, your final examination will involve a public presentation of your dissertation followed by a meeting of your committee (although University policy dictates that any graduate faculty member may attend); therefore, select a time when all members of your examination committee can allocate a 2.5-hour block. When the time and location of the defense is scheduled, your major professor will announce the meeting to HDFS faculty and graduate students. The exam will begin with a professional-level

Revised 10/2018

Procedural Guide for Ph.D. Students Human Development and Family Studies

Auburn University

presentation of the dissertation to the HDFS faculty and graduate student body. This presentation (approximately 50 minutes) is followed by a question and answer period (approximately 10 minutes) during which you will field questions from the audience about your study. The audience is then dismissed and the second part of the exam proceeds. This second part of the exam usually focuses on your dissertation study, but any aspect of your major or minor area is legitimate examination material.

The representative of the Graduate School (the university reader) also attends the defense and participates, but does not vote on your performance or sign your dissertation approval form. Only oral examinations receiving unanimous approval of the committee are successful.

If the final examination is failed, a re-examination may be given on recommendation of the advisory committee and approval by the Dean of the Graduate School.

In addition to successful completion of all examinations, the oral and written approval forms must be completed, and the final dissertation must be uploaded (i.e., published) to the Auburn University Electronic Thesis and Dissertation (AUETD) website before the degree is conferred.

Revised 10/2018

Ulasan Elemen Reka Bentuk Antara Muka bagi Meningkatkan

Keterlibatan Pelajar terhadap E-Pembelajaran di Institusi Pengajian

(Reviews on Elements of Interface Design to Improve Student Engagement towards E-Learning in

Educational Institutes)

Hazwani Nordin & Dalbir Singh

Universiti Kebangsaan Malaysia (UKM)

Fakulti Teknologi dan Sains Maklumat,

Universiti Kebangsaan Malaysia, 43600 UKM Bangi, Selangor Darul Ehsan

No. Telefon: 012-9375730, No. Faks: 03-89256732

waninordin@gmail.com, dalbir@ukm.edu.my

Abstrak

E-pembelajaran telah lama diimplementasikan dalam sistem pendidikan di Malaysia untuk meningkatkan kualiti pengajaran dan pembelajaran. Perubahan sistem pendidikan ini membolehkan komunikasi secara aktif dikalangan pelajar dan tenaga pengajar. Namun, keterlibatan dan komitmen pelajar terhadap epembelajaran masih lagi ditahap yang lemah. Pelajar lebih menerima kaedah pengajaran konvensional dan keberatan untuk menggunakan e-pembelajaran. Reka bentuk antara muka memainkan peranan penting dalam menggalakkan keterlibatan dan komitmen pelajar terhadap e-pembelajaran. Reka bentuk antara muka yang lemah merupakan penyebab ketidakselesaan pelajar untuk terus menggunakan epembelajaran. Reka bentuk antara muka e-pembelajaran yang tidak memenuhi kehendak pelajar menyebabkan mereka tidak mengambil bahagian dalam e-pembelajaran walaupun ianya boleh membantu dalam pelajaran mereka. Oleh itu, kertas ulasan ini meneliti isu berkaitan dengan elemen reka bentuk antara muka yang sentiasa berubah mengikut keperluan pengguna. Isu keterlibatan pelajar turut dibincangkan dalam kertas ulasan ini bagi memahami maksud dan situasi keterlibatan pelajar yang berbeza mengikut pelbagai tempat dan keadaan. Dapatan ulasan ini boleh digunakan dalam pembangunan antara muka e-pembelajaran yang boleh meningkatkan keterlibatan dan komitmen pelajar semasa menggunakan e-pembelajaran. Akhir sekali, hasil ulasan ini diharapkan dapat menyumbang kepada peningkatan interaksi antara pengajar dan pelajar serta turut meningkatkan kualiti pengajaran dan pembelajaran.

MyJICT - Malaysian Journal of Information and Communication Technology Vol 1 2016, Issue 2 | e-ISSN 0127-7812 89

Abstract

E-learning has been implemented in the Malaysian educational system to improve the quality of teaching and learning. The changes in the educational system has allowed active communication among students and lecturers. However, student engagement and commitment towards e-learning is still poor. Students are more accepting of conventional teaching method and reluctance to use e-learning. Interface design plays an important role to encourage student engagement and commitment towards e-learning. Poor interface design causes discomfort among students to continue in using e-learning. E-learning interface design that does not meet the student’s needs may hinder student’s participation toward e-learning although it can help them in their studies. Therefore, this article review the issues related to interface design elements that changes based on the user needs. Student engagement issues are also discussed in this review article to understand the meaning and different student engagement situations which changes based on various places and environments. The findings from this review could be used in the development of e-learning interface design that would increase the student’s engagement while using e-learning. Lastly, the result of this review is hoped to contribute in enhancing interaction between student and lecturer while improving the quality of teaching and learning.

1. PENDAHULUAN

Perkembangan teknologi maklumat dan komunikasi pada hari ini memberi peluang kepada pelajar serta tenaga pengajar untuk menyampaikan aktiviti pengajaran dan pembelajaran interaktif melalui e-pembelajaran. Penggunaan e-pembelajaran dalam sistem pendidikan hari ini dapat membantu pelajar untuk memahami pengetahuan yang kompleks dengan lebih mudah dan membolehkan pelajar mencapai pembelajaran yang bermakna (Katuk, 2013) serta membolehkan pengajaran dan pembelajaran dijalankan walaupun setiap pelajar dipisahkan oleh tempat yang berlainan dan pada waktu yang berbeza (Sithole, B. D., & Onyari, 2012). Ini menyebabkan banyak institusi pengajian telah mengimplementasikan e-pembelajaran bagi menukar strategi pengajaran dan pembelajaran daripada pembelajaran dalam bilik kuliah semata-mata kepada pembelajaran bersepadu dan secara dalam talian sepenuhnya. Namun begitu, penerapan teknologi maklumat dan komunikasi dalam sistem pendidikan masa kini menyebabkan ramai pengguna mahukan reka bentuk antara muka yang berkualiti tinggi ketika berinteraksi dengan sistem e-pembelajaran. Reka bentuk antara muka merupakan teras dan komponen bersepadu bagi keseluruhan sistem e-pembelajaran (Reyna, 2009). Ini kerana reka bentuk antara muka e-pembelajaran perlu ditentukan berdasarkan bagaimana pelajar

MyJICT - Malaysian Journal of Information and Communication Technology Vol 1 2016, Issue 2 | e-ISSN 0127-7812 90

ingin belajar dan tugasan yang pelajar perlukan untuk menjalankan e-pembelajaran (Guralnick 2006; Reyna 2009).

Reka bentuk antara muka yang baik dapat meningkatkan keterlibatan pelajar terhadap e-pembelajaran (Zaharias 2008; Wang et al. 2010; Faghih et al. 2013). Ini kerana reka bentuk antara muka memainkan peranan penting dalam meningkatkan rupa dan rasa (look and feel) e-pembelajaran dan ia boleh mempengaruhi cara pelajar berinteraksi dengan e-pembelajaran (Reyna, 2009). Rupa dan rasa reka bentuk antara muka e-pembelajaran boleh mempengaruhi pelajar sama ada ingin terus berinteraksi dengan e-pembelajaran atau sebaliknya. Komunikasi yang efektif dapat meningkatkan keterlibatan dan motivasi pelajar serta menjadikan interaksi pelajar dengan e-pembelajaran menjadi lebih menarik (Wang et al., 2010). Selain itu, keterlibatan pelajar juga boleh meningkatkan kepuasan pelajar ketika berinteraksi dengan e-pembelajaran dan seterusnya membantu pelajar meningkatkan pencapaian pelajar dalam bidang akademik (Essam & Al-Ammary, 2013).

Penglibatan pelajar secara berterusan dalam e-pembelajaran penting kerana ia menjurus kepada pembelajaran yang lebih baik. Kebanyakan institusi pengajian telah mengimplementasikan e-pembelajaran tersendiri bagi memenuhi keperluan pengajaran dan pembelajaran pelajar. Selain daripada e-pembelajaran yang dibangunkan oleh pihak institusi, terdapat e-pembelajaran yang ditawarkan secara percuma kepada pelajar seperti ‘google classroom’ dan ‘open learning’. Tenaga pengajar juga memainkan peranan dalam memastikan keterlibatan pelajar dalam e-pembelajaran dengan memberi markah kepada pelajar yang menyertai perbincangan di forum, dan memberikan latihan secara berterusan secara melalui e-pembelajaran. Namun begitu, aspek reka bentuk antara muka e-pembelajaran tidak boleh diabaikan kerana antara muka merupakan titik permulaan anggapan pelajar terhadap e-pembelajaran di mana kejayaan atau kegagalan e-pembelajaran bergantung kepada reka bentuk antara muka (Faghih, Azadehfar & Katebi 2013). Tambahan pula, reka bentuk antara muka e-pembelajaran adalah sangat kritikal kerana keberkesanan pembelajaran dan reka bentuk antara muka mempunyai perkaitan yang sangat ketara (Guralnick, 2006). Oleh itu, kertas ulasan ini menumpukan kepada reka bentuk antara muka untuk meningkatkan keterlibataan pelajar dalam e-pembelajaran di institusi pengajian.

2. KETERLIBATAN PELAJAR TERHADAP E-PEMBELAJARAN

Istilah keterlibatan pelajar (student engagement) semakin banyak diguna pakai dalam bidang pendidikan sejak beberapa tahun kebelakangan ini. Keterlibatan pelajar dilihat sebagai petunjuk kepada kejayaan pengajaran dan pembelajaran pelajar dalam e-pembelajaran dan ia sangat dihargai sebagai sebahagian daripada aktiviti penambahbaikan dalam bidang akademik. Antara faktor utama keterlibatan pelajar penting dalam e-pembelajaran adalah disebabkan oleh kadar keciciran pelajar yang tinggi dalam e-pembelajaran di universiti (Angelino, Williams & Natvig 2007; Backs 2012). Selain itu, pendidik turut memberi penekanan bagi meningkatkan keterlibatan pelajar dalam e

MyJICT - Malaysian Journal of Information and Communication Technology Vol 1 2016, Issue 2 | e-ISSN 0127-7812 91

pembelajaran sebagai jalan untuk menangani pencapaian yang rendah, kebosanan ketika sesi pembelajaran dan pengasingan pelajar (Fredricks et al. 2011).

Beberapa penyelidik terdahulu telah mengkaji mengenai keterlibatan pelajar dalam e-pembelajaran dan telah mengemukakan pendapat dan definisi yang berbeza. Oleh itu, memahami takrifan keterlibatan pelajar dengan lebih mendalam adalah penting walaupun ramai penyelidik mengemukakan definisi yang berlainan namun ia mampu membantu kajian ini bagi memahami maksud keterlibatan pelajar dengan lebih mendalam.

3. DEFINISI KETERLIBATAN PELAJAR

Istilah keterlibatan pelajar pertama kalinya muncul pada akhir tahun 1980-an. Ketika itu, kebanyakan penyelidik cenderung untuk melihat sebab musabab keterlibatan pelajar melalui satu set faktor di mana demografi dan latar belakang sosial dikaitkan dengan pelajar seperti keadaan keluarga dan pengaruh rakan sebaya. Namun, konsep keterlibatan pelajar mula beralih makna hasil daripada peningkatan perhatian dan pengaruh sekolah terutamanya hubungan antara institusi pendidikan dengan pengalaman keterlibatan pelajar (Dunleavy & Milton, 2009). Keterlibatan pelajar mempunyai pelbagai maksud kepada ramai penyelidik dalam kajian yang berbeza serta ia boleh ditakrifkan secara berbeza mengikut tempat. Sebagai contoh, konsep keterlibatan pada sesebuah sekolah boleh diperhatikan seperti menghadiri kelas, perubahan kerja dalam jangka masa yang diberikan dan mengikut peraturan sekolah. Sementara itu, di sekolah lain konsep keterlibatan difahami dari segi dalaman seperti semangat, rasa ingin tahu, keyakinan, motivasi dan faedah (Ching, 2012).

Istilah keterlibatan memberi maksud lebih daripada penglibatan di mana ia memerlukan perasaan dan membuat pertimbangan serta aktiviti (Trowler, 2010). Menurut Trowler (2010) penglibatan tanpa perasaan diibaratkan seperti berlakon tanpa perasaan dan hanya terlibat untuk patuh kepada skrip. Welch, & Bonnan-White (2012) menyatakan bahawa keterlibatan pelajar bermaksud masa dan usaha pelajar yang tertumpu kepada aktiviti-aktiviti yang secara empirikal dikaitkan dengan hasil yang dikehendaki oleh institusi dan apa yang dilakukan oleh institusi untuk mendorong pelajar menyertai aktiviti-aktiviti ini seperti belajar, berinteraksi dengan pengajar dan rakan sebaya, melibatkan diri dalam aktiviti kampus dan memenuhi jangkaan akademik yang tinggi.

Dunleavy, & Milton (2009) menyatakan bahawa definisi keterlibatan pelajar adalah berbeza dan bergantung kepada pendekatan yang digunakan. Beliau menyatakan bahawa keterlibatan pelajar melibatkan dua dimensi iaitu keterlibatan sosial dan keterlibatan akademik. Beliau menakrifkan keterlibatan sosial sebagai gabungan semangat kekitaan di kalangan pelajar di institusi pengajian, penerimaan mereka terhadap matlamat institusi, perasaan yang berhubung dan diterima rakan-rakan dan pengalaman berhubungan dengan orang dewasa yang menunjukkan minat mereka sebagai individu. Pada awal tahun 1990-an, Newmann, Wehlage & Lamborn (1992) menakrifkan keterlibatan akademik sebagai pelaburan psikologi pelajar dan usaha yang menjurus kepada pembelajaran,

MyJICT - Malaysian Journal of Information and Communication Technology Vol 1 2016, Issue 2 | e-ISSN 0127-7812 92

pemahaman atau penguasaan pengetahuan dan kemahiran bagi menggalakkan prestasi akademik. Beliau turut meneruskan kajian bagi memahami apa yang mendorong pelajar untuk berjaya dalam akademik seperti menyiapkan latihan dan kerja sekolah, menghadiri kelas serta mengambil bahagian di dalam bilik darjah (Dunleavy, & Milton 2009; Ching 2012). Selain itu, ramai penyelidik lain yang berpendapat bahawa takrifan keterlibatan pelajar adalah bergantung kepada dimensi yang berbeza seperti tingkah laku (behavioural), emosi (emotional) dan kognitif (cognitive).

Terdapat banyak definisi mengenai keterlibatan pelajar namun tiada definisi yang khusus mengenainya. Definisi yang dikemukakan oleh penyelidik lepas adalah berdasarkan kepada kajian yang dijalankan oleh mereka pada masa itu di mana ia dirasakan bersesuaian dengan situasi kajian mereka pada masa tersebut. Walaupun definisi yang dikemukakan berbeza antara satu sama lain tetapi perbezaan tersebut tidak terlalu ketara dan hampir kesemua definisi mempunyai matlamat yang sama iaitu menjurus kepada penambahbaikan dalam pendidikan pelajar. Oleh itu, definisi keterlibatan pelajar dalam kajian ini menjurus kepada menjalankan atau melakukan aktiviti yang disertakan dengan perasaan dalaman yang tinggi yang mana ia boleh mendorong pelajar ke arah yang lebih baik. Oleh kerana kebanyakan definisi yang dikemukakan melibatkan tugasan atau aktiviti maka definisi yang sedia ada telah menepati kehendak kajian.

4. FAKTOR YANG MEMPENGARUHI KETERLIBATAN PELAJAR

Kejayaan dan kegagalan e-pembelajaran bergantung kepada keterlibatan pelajar. Berdasarkan kajian lepas beberapa faktor yang boleh mempengaruhi keterlibatan pelajar dalam e-pembelajaran telah dikenal pasti. Antara faktor tersebut adalah seperti berikut:

a) Kandungan kursus

Di kebanyakan institusi pengajian, kandungan kursus disediakan oleh pengajar. Ladbrook, & Parr (2015) menyatakan bahawa kandungan kursus menyumbang kepada penglibatan pelajar secara aktif dalam e-pembelajaran. Beliau menyarankan agar kandungan kursus diberi perhatian dari segi reka bentuk kursus kerana ia mampu meningkatkan motivasi dan penglibatan pelajar dalam e-pembelajaran. Namun, ramai pelajar yang menarik diri daripada kursus yang diikuti dalam e-pembelajaran berbanding dengan pembelajaran dalam bilik kuliah (Angelino et al. 2007; Ali, & Smith 2015; Boton et al. 2015). Menurut Boton et al. (2015), sebab utama pelajar menarik diri daripada kursus berkenaan kerana kandungan kursus yang ditawarkan tidak sesuai dengan keperluan pelajar. Selain itu, tugasan yang diberikan tidak mencabar menyebabkan pelajar lebih cepat bosan (Dixson, 2010) dan tidak merasa cabaran ketika menyiapkan tugasan tersebut. Oleh itu, mereka bentuk kursus adalah tanggungjawab pengajar dalam meningkatkan keterlibatan pelajar dan motivasi pelajar untuk menggunakan e-pembelajaran sebagai medium pembelajaran mereka.

MyJICT - Malaysian Journal of Information and Communication Technology Vol 1 2016, Issue 2 | e-ISSN 0127-7812 93

b) Alatan yang ditawarkan

Alatan (tools) yang ditawarkan dalam e-pembelajaran turut mempengaruhi motivasi dan keterlibatan pelajar (Delialioǧlu 2012; Ladbrook, & Parr 2015). Alatan yang dimaksudkan seperti audio, visual dan teks adalah penting dalam membolehkan sesuatu komunikasi berlaku dalam e-pembelajaran. Audio dan visual yang berkualiti rendah boleh mengganggu komunikasi pelajar ketika sesi perbincangan dan secara tidak langsung boleh mempengaruhi motivasi dan keterlibatan pelajar untuk terus menggunakannya. Selain itu, teks yang digunakan mesti bersesuaian dari segi jenis tulisan dan saiz teks dengan cita rasa pengguna seperti pengguna yang berusia, remaja mahupun pelajar sekolah.

c) Budaya Pelajar

Budaya sering dikaitkan dengan keterlibatan pelajar dalam e-pembelajaran. Setiap pelajar terdiri daripada latar belakang budaya yang berbeza (Chakrabarty & Mohamed, 2013) namun perbezaan budaya ini tidak bermaksud malas atau tidak berusaha dalam pembelajaran tetapi ia bermaksud setiap pelajar mempunyai keutamaan yang berbeza dalam persekitaran pembelajaran mereka (Parsons & Taylor, 2011). Budaya terdiri daripada kepercayaan, andaian, pengetahuan, nilai-nilai atau amalan yang dikongsi dalam masyarakat (Benaida, 2014; Mcfarlane, 2011). Sebagai contoh, terdapat sesetengah budaya yang mana pelajar lebih menerima pembelajaran konvensional berbanding pembelajaran menggunakan e-pembelajaran (Zhang, 2013). Oleh itu, faktor budaya tidak boleh dipandang ringan dalam usaha meningkatkan keterlibatan pelajar menggunakan e-pembelajaran di institusi pengajian.

d) Reka bentuk antara muka

Reka bentuk antara muka pengguna tidak boleh dipandang remeh bagi meningkatkan keterlibatan pelajar dalam e-pembelajaran. Ini kerana reka bentuk antara muka pengguna adalah bahagian yang paling penting dan kritikal dalam e-pembelajaran kerana keberkesanan pembelajaran dan reka bentuk antara muka mempunyai perkaitan dan saling bergantung antara satu sama lain (Guralnick, 2006). Pelajar pada hari ini yang sentiasa terdedah dengan teknologi terkini berasa cepat bosan ketika menggunakan e-pembelajaran yang mempunyai antara muka yang tidak menarik (Parsons & Taylor, 2011) seperti antara muka yang mempunyai banyak teks, visual yang membosankan serta kesukaran mencari maklumat. Reka bentuk antara muka e-pembelajaran hendaklah ditentukan oleh bagaimana pelajar belajar dan bagaimana mereka menggunakan e-pembelajaran untuk melaksanakan sebarang tugasan (Reyna, 2009). Oleh itu, reka bentuk susun atur memainkan peranan penting dalam rupa dan rasa e-pembelajaran dan boleh mempengaruhi cara pelajar belajar dan berinteraksi menggunakan e-pembelajaran.

MyJICT - Malaysian Journal of Information and Communication Technology Vol 1 2016, Issue 2 | e-ISSN 0127-7812 94

5. DEFINISI REKA BENTUK ANTARA MUKA

Aplikasi teknologi hari ini adalah berpaksi kepada pembangunan antara muka kerana reka bentuk antara muka yang baik dapat melancarkan perjalanan tugasan tetapi berurusan dengan antara muka yang rumit dan mengelirukan boleh menyebabkan pengguna mengelak daripada menggunakan sistem tersebut (Barzegar, Shojafar, & Keyvanpour, 2009). Reka bentuk antara muka hendaklah dibangunkan dengan teliti dan menepati kehendak dan piawaian pengguna agar pengguna dapat berinteraksi dengan aplikasi sistem semaksima yang mungkin.

Pelbagai takrifan antara muka yang telah dikemukakan oleh penyelidik lepas. Galitz (2007) menyatakan antara muka adalah sebahagian daripada komputer dan perisiannya yang membolehkan pengguna mendengar, melihat, menyentuh, bercakap dan memahami secara langsung atau tidak langsung. Beliau mengkategorikan antara muka kepada dua komponen iaitu input dan output. Input adalah bagaimana pengguna berkomunikasi dengan komputer menggunakan komponen input seperti papan kekunci, tetikus, skrin sentuh dan suara. Output pula adalah bagaimana komputer menyampaikan hasil dan keperluan kepada pengguna.

Adnan, Ali & Ahmad (2015) menakrifkan antara muka sebagai satu alat atau program yang membolehkan pengguna berkomunikasi dengan sistem komputer. Antara muka didefinisikan sebagai medium perantaraan bagi komunikasi antara manusia dengan sistem komputer. Antara muka digunakan untuk menghantar maklumat daripada manusia untuk sistem yang dikenali sebagai interaksi. Dalam teknologi maklumat dan komunikasi (ICT), reka bentuk antara muka merujuk kepada permukaan skrin sesuatu alat seperti komputer, telefon pintar dan tablet yang membolehkan pengguna untuk mengakses dan memahami maklumat yang dikehendaki (Kamaruddin, Park & Nam 2012). Menurut Kamaruddin (2014), reka bentuk antara muka tidak hanya berkaitan dengan nilai-nilai estetika, rupa dan rasa yang dibentuk oleh elemen-elemen reka bentuk seperti warna, teks, imej dan susun atur tetapi merupakan medium komunikasi yang menjadi perantara antara pengguna dan sistem.

Pelbagai definisi antara muka telah dikemukakan oleh penyelidik lepas, namun kebanyakan kajian menyatakan bahawa antara muka adalah medium komunikasi antara manusia dan sistem yang diterjemahkan dalam bentuk permukaan skrin. Oleh itu, dalam kajian ini antara muka merujuk kepada permukaan skrin komputer bagi memudahkan interaksi dengan manusia.

6. KAEDAH DAN PROSEDUR PEMILIHAN ELEMEN ANTARA MUKA

Kajian ini adalah untuk menentukan elemen antara muka yang boleh mempengaruhi keterlibatan pelajar dalam e-pembelajaran. Jadual 1 di bawah menunjukkan kaedah yang digunakan dalam pemilihan elemen antara muka e-pembelajaran.

MyJICT - Malaysian Journal of Information and Communication Technology Vol 1 2016, Issue 2 | e-ISSN 0127-7812 95

Jadual 1 Pemilihan Prosedur

Prosedur Penerangan mengenai prosedur

Pemilihan artikel Mencari artikel yang berkaitan dengan reka bentuk antara muka.

Pemilihan elemen antara muka Memilih elemen antara muka yang sering digunakan dalam e-pembelajaran.

Perincikan elemen antara muka Memperincikan ciri-ciri setiap elemen antara muka yang telah dipilih.

6.1 Pemilihan Artikel

Bagi menentukan elemen antara muka yang boleh mempengaruhi keterlibatan pelajar, maklumat yang diperolehi daripada artikel, prosiding dan buku rujukan dikaji dengan lebih mendalam. Antara kriteria-kriteria yang digunakan ketika proses pemilihan artikel adalah seperti berikut:

• Artikel yang diterbitkan dari tahun 2000 hingga 2016 diperolehi daripada pangkalan data dalam talian. Jangka masa lebih dari satu dekad dipilih adalah untuk melihat sama ada terdapat perubahan elemen-elemen antara muka e-pembelajaran selama sedekad ini.

• Elemen reka bentuk antara muka: Artikel-artikel yang dikaji memfokuskan kepada elemen reka bentuk antara muka sahaja. Elemen antara muka yang boleh mempengaruhi pelajar untuk terus terlibat dalam menggunakan e-pembelajaran.

• Tempatan dan antarabangsa: Elemen antara muka e-pembelajaran dikaji secara meluas iaitu melibatkan artikel tempatan dan luar negara. Ini bertujuan untuk mengenal pasti persamaan dan perbezaan pemilihan elemen-elemen antara muka bagi dalam dan luar negara.

• Kata kunci: beberapa kata kunci digunakan semasa proses pencarian artikel bagi mendapatkan artikel yang berkaitan sahaja. Antara kata kunci yang digunakan adalah elemen antara muka, antara muka pengguna, keterlibatan pelajar, corak antara muka dan pilihan antara muka.

6.2 Pemilihan Elemen Antara Muka

Semua artikel yang berkaitan dikaji dan dianalisis dengan teliti bagi mengenal pasti apa fungsi setiap elemen dan ciri-ciri antara muka yang mendorong kepada kejayaan e-pembelajaran. Antara langkah-langkah yang diambil dalam memilih elemen antara muka adalah seperti berikut:

• Kenal pasti elemen antara muka: Menyenaraikan semua elemen antara muka yang dikaji oleh penyelidik lepas. Setiap fungsi elemen tersebut juga dikenal pasti.

• Kepentingan elemen antara muka: Setiap elemen antara muka dikenal pasti mesti sesuai kepada pengguna sasaran jika ia diimplemenkan ke dalam antara muka e-pembelajaran.

MyJICT - Malaysian Journal of Information and Communication Technology Vol 1 2016, Issue 2 | e-ISSN 0127-7812 96

• Kurangkan elemen yang tidak penting: Penggunaan elemen antara muka yang terlalu banyak boleh menyebabkan antara muka nampak terlalu sesak sehingga mengaburi objektif utama e-pembelajaran dibangunkan.

6.3 Perincian Elemen Antara Muka

Elemen antara muka yang telah dipilih mesti dikaji dengan lebih mendalam bagi memastikan ia memenuhi kehendak dan keperluan pengguna. Setiap elemen juga perlu dikenal pasti kefungsiannya. Perincian elemen antara muka juga dilaksanakan dengan mengadaptasikan ke dalam prinsip reka bentuk antara muka bagi membangunkan antara muka yang baik.

7. ELEMEN ANTARA MUKA MEMPENGARUHI KETERLIBATAN PELAJAR DALAM E-PEMBELAJARAN

Elemen antara muka penting dalam mereka bentuk antara muka. Elemen antara muka ditakrifkan sebagai sesuatu yang boleh dilihat (seperti gambar, carta atau animasi) yang digunakan untuk menjadikan antara muka tersebut lebih menarik dan mudah untuk difahami (Adnan et al., 2015). Masalah utama dalam e-pembelajaran adalah apabila pensyarah dan pembangun aplikasi lebih mengutamakan reka bentuk kursus dan tidak memberi perhatian kepada reka bentuk antara muka e-pembelajaran (Thomson, 2014). Akibatnya, pelajar hanya dapat melihat terlalu banyak teks, imej, pautan, grafik dan kurang keharmonian dalam reka bentuk antara muka e-pembelajaran. Ini menyebabkan pelajar sukar untuk memberi tumpuan, memahami dan belajar dengan cara yang diingin. Oleh itu, pelajar menjadi kurang berminat dengan pembelajaran menggunakan e-pembelajaran.

Elemen antara muka boleh memberi kesan ke atas rupa dan rasa reka bentuk antara muka e-pembelajaran dan sekaligus boleh mempengaruhi keterlibatan pelajar terhadap e-pembelajaran. Banyak elemen reka bentuk antara muka yang dikemukakan oleh penyelidik lepas dikategorikan kepada warna (Adnan et al., 2015; Arockiam & Selvaraj, 2013; Bakar & Long, 2013; Faghih et al., 2013; Khanum, Fatima, & Chaurasia, 2012; Reyna, 2009, 2013; Tomczyk, 2009), imej (Adnan et al., 2015; Faghih et al., 2013; Khanum et al., 2012), teks (Adnan et al., 2015; Arockiam & Selvaraj, 2013; Bakar & Long, 2013; Khanum et al., 2012; Reyna, 2009, 2013), susun atur (Khanum et al., 2012; Reyna, 2013), navigasi (Reyna, 2009), animasi (Faghih et al., 2013), bentuk (Adnan et al., 2015) dan pautan (Khanum et al., 2012). Jadual 2 menunjukkan kajian yang dijalankan oleh penyelidik lepas dan kekerapan elemen reka bentuk antara muka yang digunakan dalam kajian terdahulu.

MyJICT - Malaysian Journal of Information and Communication Technology Vol 1 2016, Issue 2 | e-ISSN 0127-7812 97

Jadual 2 Elemen antara muka pada kajian lepas

Reyna (2009) X X X X

Tomczyk (2009) X X

Kamaruddin (2012) X X X X

Ramakrisnan, Jaafar, X X X

Razak (2012)

Bakar & Long (2013) X X

Faghih et al. (2013) X X

Shahzad, Hafizi, & X

Golamdin (2014)

Yaa & Adzobu X X (2014)

Adnan et al. (2015) X X X

Rajah 1 merupakan hasil daripada sorotan susastera yang dijalankan yang mana ramai penyelidik lepas berpendapat bahawa warna, tipografi, imej, susun atur dan navigasi memainkan peranan penting dalam reka bentuk antara bagi meningkatkan keterlibatan pelajar dalam e-pembelajaran. Kajian yang dijalankan oleh Reyna (2009) mengkategorikan elemen ikon, simbol dan banner sebagai imej. Maka, elemen ikon, simbol dan banner dikategori dalam elemen imej untuk kajian ini. Kertas ulasan ini hanya berkisarkan mengenai elemen antara muka yang memberi kesan ke atas keterlibatan pelajar menggunakan e-pembelajaran. Berikut adalah antara elemen yang telah dikenal pasti oleh penyelidik lepas yang dikupas dengan lebih mendalam bagi meningkatkan keterlibatan pelajar menggunakan e-pembelajaran di institusi pengajian.

MyJICT - Malaysian Journal of Information and Communication Technology Vol 1 2016, Issue 2 | e-ISSN 0127-7812 98

Rajah 1: Elemen Reka Bentuk Antara Muka E-Pembelajaran

a) Warna

Warna menjadi elemen yang penting dalam antara muka e-pembelajaran. Ini kerana warna berperanan dalam menarik minat pelajar dan menjana keinginan untuk terus menggunakan e-pembelajaran. Terdapat beberapa warna yang boleh meningkatkan tindak balas pelajar dan turut memberi kesan yang menenangkan kepada pengguna (Thomson, 2014). Menurut Müller (2014), warna bukan hanya sekadar warna tetapi boleh mewujudkan emosi, perasaan (Reyna, 2013) dan kenangan yang boleh mempengaruhi minda dan badan manusia melalui fizikal dan juga reaksi psikologi kepada warna tertentu. Kombinasi warna yang betul boleh menekankan organisasi skrin, memudahkan diskriminasi komponen skrin, menyerlahkan perbezaan dan dapat membuatkan paparan lebih menarik. Jika warna tidak digunakan dengan betul boleh menyebabkan keletihan visual, pening, sakit mata, sakit kepala, hilang tumpuan atau motivasi (Reyna, 2009) dan boleh merosakkan kebolehgunaan sistem (Galitz, 2007). Jadual 3 menunjukkan kajian yang dijalankan oleh Galitz (2007) mengenai kombinasi warna latar belakang dan latar depan yang sesuai.Berikut adalah senarai kombinasi warna latar belakang (background) dengan warna latar depan (foreground) yang sesuai seperti dalam Jadual 3.

Jadual 3 Contoh kombinasi warna yang sesuai

Warna Latar Belakang (Background) Warna Latar Depan (Foreground)

Hitam Cyan gelap Cyan terang

Kuning gelap Kuning terang

Putih Magenta terang Hijau terang

Biru Hijau gelap Hijau terang

MyJICT - Malaysian Journal of Information and Communication Technology Vol 1 2016, Issue 2 | e-ISSN 0127-7812 99

Kuning gelap Cyan terang

Putih Kuning terang

Hijau Hitam Kuning terang

Biru gelap Putih

Merah Hijau terang Kuning terang

Cyan terang Putih

Kuning Hitam Merah gelap

Biru gelap

Putih Hitam

Biru gelap

Cyan Hitam Kuning terang

Biru gelap Putih

Magenta Hitam Kuning

Cyan terang Putih

Selain itu, penggunaan warna dalam e-pembelajaran turut bergantung kepada pengguna. Jika pengguna e-pembelajaran adalah kanak-kanak, maka warna terang adalah paling sesuai digunakan. Kajian yang dijalankan oleh Adnan et al. (2015) menyatakan bahawa warna biru adalah warna yang paling popular dikalangan kanak-kanak. Sementara itu, kajian yang dijalankan oleh Müller (2014) ke atas pelajar di universiti di Afrika Selatan menyatakan bahawa warna biru, merah, ungu dan hitam adalah warna pilihan pelajar di universiti tersebut.

Selain daripada warna latar belakang, penekanan warna grafik, teks, butang menu dan pautan juga harus diberi perhatian. Ini kerana kesilapan terhadap kombinasi warna yang sesuai boleh mengundang kepada penurunan penggunaan e-pembelajaran dikalangan pelajar. Tambahan pula, terdapat warna yang boleh mengganggu keterlibatan pelajar dan tidak sesuai untuk digunakan dalam e-pembelajaran (Thomson, 2014). Ini kerana ia bertentangan dengan budaya sesuatu masyarakat. Menurut Kim, & Kuljis (2010) warna mempunyai maksud yang berlainan bagi budaya yang berlainan. Sebagai contoh, di Jepun warna putih melambangkan kematian manakala di Egypt melambangkan kegembiraan. Oleh itu, penggunaan warna perlu dipilih dengan teliti dan cermat agar tidak menyinggung mana-mana budaya.

b) Imej, grafik, logo dan ikon

Elemen imej sering diperkatakan oleh penyelidik lepas sebagai salah satu elemen yang mampu mempengaruhi keterlibatan pelajar dalam e-pembelajaran. Penggunaan imej dalam antara muka e-pembelajaran dapat membantu pelajar memahami isi kandungan tanpa perlu membaca teks. Imej pada

MyJICT - Malaysian Journal of Information and Communication Technology Vol 1 2016, Issue 2 | e-ISSN 0127-7812 100

antara muka terdiri daripada ikon, simbol dan gambar (Adnan et al., 2015). Setiap imej yang digunakan mempunyai maksud yang tersendiri yang difahami oleh para pengguna tanpa perlu membaca teks atau label. Penggunaan imej yang sesuai serta menarik boleh memikat lebih ramai pelajar untuk belajar menggunakan e-pembelajaran, memudahkan pembelajaran dan menguatkan ingatan (Adnan et al., 2015). Ini kerana, imej lebih mudah untuk dikenal pasti dan diingati berbanding dengan penggunaan teks semata-mata.

Walau bagaimanapun, penggunaan imej yang banyak juga boleh mempengaruhi minat pelajar untuk menggunakan e-pembelajaran. Imej yang banyak dalam e-pembelajaran boleh mengelirukan pelajar walaupun imej yang digunakan berkaitan dengan kursus yang diambil serta boleh menyebabkan pelajar hilang fokus terhadap apa yang ingin mereka pelajari daripada e-pembelajaran (Thomson 2014; Adnan et al. 2015). Selain itu, saiz imej juga harus diberi perhatian ketika membangunkan antara muka e-pembelajaran. Ini kerana saiz imej yang terlalu besar atau terlalu kecil boleh menjadi isu pelajar tidak mahu menggunakan e-pembelajaran. Namun, terdapat pelajar yang tidak terganggu dengan bilangan dan saiz imej dalam antara muka e-pembelajaran. Menurut kajian yang dijalankan oleh Kim, & Kuljis (2010), negara Korea Selatan lebih gemar jika antara muka mempunyai banyak imej berbanding dengan United Kingdom yang lebih gemar jika penggunaan imej adalah minima. Perbezaan ini dikaitkan dengan latar belakang budaya yang berbeza yang menyebabkan kehendak setiap pelajar adalah berbeza.

Bagi meningkatkan keterlibatan pelajar, imej yang diletakkan pada e-pembelajaran perlu diminimumkan dan berstruktur. Selain itu, imej pada antara muka hendaklah dalam keadaan bersusun bagi memudahkan pelajar mencari maklumat dan tidak sesat ketika melayari e-pembelajaran (Galitz, 2007). Menurut Thomson (2014) meminimakan penggunaan grafik selari dengan teks adalah salah satu cara untuk mengekalkan minat, motivasi dan keterlibatan pelajar terhadap e-pembelajaran. Oleh itu, ruang putih yang banyak pada antara muka e-pembelajaran dibenarkan bagi membolehkan pelajar lebih fokus kepada satu-satu tugasan.

c) Teks / Tipografi

Elemen tipografi penting dalam reka bentuk antara muka e-pembelajaran sebagai medium untuk menyampaikan maklumat. Ia berfungsi untuk menghuraikan visual seperti label untuk ikon yang mana menjadi pilihan apabila simbol atau ikon tidak dapat menyampaikan mesej yang ingin disampaikan (Faghih et al. 2013; Adnan et al. 2015). Tambahan pula, tipografi merangkumi jenis dan saiz teks yang mempunyai kesan ke atas kebolehgunaan sesuatu sistem (Khanum et al., 2012). Teks yang digunakan pada antara muka bukan hanya menulis kandungan tetapi dimanipulasikan sebagai pautan, butang dan animasi yang berperanan dalam navigasi, interaksi dan penyampaian e-pembelajaran (Adnan et al., 2015). Namun, penggunaan teks semata-mata dalam antara muka membuatkan antara muka menjadi kurang menarik (Zhang 2013; Faghih et al. 2013). Tambahan pula,

MyJICT - Malaysian Journal of Information and Communication Technology Vol 1 2016, Issue 2 | e-ISSN 0127-7812 101

dalam membantu proses pembelajaran pelajar penyediaan antara muka berasaskan teks sahaja tidak mencukupi untuk menjadikan pembelajaran berkesan, cekap dan membuatkan pelajar terus melibatkan diri (Bart, 2012). Penggunaan teks yang terlalu banyak dalam antara muka boleh mengelirukan pelajar dan mereka tidak dapat membezakan apa yang relevan dengan pembelajaran mereka sama ada penting atau tidak (Thomson, 2014) kerana kekurangan penstrukturan teks boleh menjejaskan kepentingan minat pelajar dalam menggunakan e-pembelajaran. Tambahan pula, bentuk atau jenis fon dalam e-pembelajaran perlu dipilih dengan teliti. Fon yang digunakan hendaklah sesuai dengan tujuan e-pembelajaran diwujudkan dan tidak terlalu mencolok dan mewah (Thomson, 2014) sehingga mampu mengalih perhatian pelajar ketika proses pembelajaran berlangsung. Fon yang digunakan perlu membuatkan antara muka nampak estetik bagi meningkatkan keterlibatan pelajar dalam e-pembelajaran. Selain itu, penggunaan fon mestilah sesuai dengan peringkat umur pelajar sasaran kerana jenis fon bagi yang digemari bagi kanak-kanak, orang dewasa serta orang tua adalah tidak sama antara satu sama lain (Tomczyk, 2009).

Selain itu, penggunaan warna teks juga penting dalam menarik lebih ramai pelajar untuk menggunakan e-pembelajaran. Penggunaan warna teks dan warna latar belakang perlu kontra seperti warna latar belakang yang cerah dan warna teks yang gelap agar pengguna mudah untuk membaca teks. Selain itu, saiz teks juga perlu diberi perhatian kerana teks yang terlalu kecil atau besar boleh menjejaskan tumpuan pelajar ketika menggunakan e-pembelajaran. Oleh itu, pembangun antara muka e-pembelajaran perlu mengetahui kehendak dan keperluan pelajar dalam sesebuah antara muka e-pembelajaran agar pembelajaran dapat berjalan dengan sempurna dan lancar.

d) Susun atur (Layout)

Susun atur antara muka merupakan elemen yang paling penting kerana pengguna memberikan perhatian ketika berkomunikasi dengan e-pembelajaran. Susun atur ditakrifkan sebagai bagaimana elemen-elemen antara muka seperti imej, butang, pautan, logo, skim warna dan seni bina maklumat diagihkan pada halaman antara muka (Tomczyk 2009; Reyna 2013). Ramai penyelidik lepas yang menyatakan bahawa susun atur fizikal objek visual pada skrin memainkan peranan dalam persepsi estetika pengguna terhadap antara muka (Altaboli & Lin, 2011) dan susun atur antara muka penting untuk mendapatkan kepercayaan daripada pengguna ketika menggunakan e-pembelajaran (Alias, Zakariah, Ismail, & Aziz, 2012).

Susun atur yang konsisten amat penting dalam antara muka e-pembelajaran kerana ia melibatkan kredibiliti dan pengajaran serta pembelajaran secara professional. Susun atur e-pembelajaran yang konsisten mempunyai halaman dengan susun atur bentuk dan corak yang sama di semua halaman e-pembelajaran (Tomczyk 2009; Reyna 2013). Susun atur yang tidak konsisten memberi kesan terhadap navigasi pengguna (Lim, Ayesh, & Chee, 2013). Sebagai contoh, butang menu diletakkan di bahagian atas pada halaman pertama e-pembelajaran tetapi butang menu

MyJICT - Malaysian Journal of Information and Communication Technology Vol 1 2016, Issue 2 | e-ISSN 0127-7812 102

diletakkan pada sebelah tepi pada halaman berikutnya. Ini menyebabkan pengguna menghabiskan banyak masa mencari butang yang sama pada halaman berlainan.

e) Navigasi

Navigasi adalah elemen penting yang perlu diambil kira ketika membangunkan antara muka e-pembelajaran. Navigasi penting bagi membolehkan pergerakan yang lancar dalam halaman e-pembelajaran. Navigasi yang baik akan membolehkan pelajar terus terlibat dalam mengendalikan tugas ketika menggunakan e-pembelajaran (Kamaruddin, 2010) dan dapat menggalakkan pelajar untuk menerokai e-pembelajaran secara keseluruhan manakala navigasi yang tidak dibangunkan dengan baik akan menyebabkan pelajar terperangkap dan bergerak di ruang yang sama (Reyna, 2009). Ini menyebabkan pelajar akan mengambil masa yang lama dan berasa kecewa ketika cuba berinteraksi menggunakan e-pembelajaran.

Kekurangan navigasi yang lancar menyukarkan para pelajar untuk menggunakan e-pembelajaran. Navigasi yang lancar adalah berpandukan kepada bantuan navigasi seperti anak panah, ikon dan butang yang perlu direka dengan baik, mudah dilihat dan senang difahami dan dilayari (Thomson, 2014). Walau bagaimanapun, antara muka e-pembelajaran yang mengandungi terlalu banyak bantuan navigasi akan menyukarkan pelajar dalam mengenali bentuk bantuan yang ditawarkan. Oleh itu, navigasi dalam antara muka e-pembelajaran adalah penting agar pelajar mudah untuk mencari maklumat dan menggunakan e-pembelajaran. Selain itu, navigasi yang lancar dapat memberi motivasi serta meningkatkan keterlibatan pelajar meneroka dan menggunakan e-pembelajaran.

8. PERBINCANGAN: HUBUNG KAIT REKA BENTUK ANTARA MUKA E

PEMBELAJARAN DAN KETERLIBATAN PELAJAR

Reka bentuk antara muka dan keterlibatan pelajar merupakan dua aspek yang berbeza tetapi ia saling berkaitan antara satu sama lain kerana keberkesanan antara muka e-pembelajaran ditentu berdasarkan penglibatan pelajar. Ulasan di atas telah membincangkan mengenai elemen antara muka yang mempengaruhi keterlibatan pelajar dalam e-pembelajaran yang diimplementasikan di institusi pengajian. Oleh itu, kertas ulasan ini adalah untuk mengulas mengenai elemen reka bentuk antara muka yang memainkan peranan dalam memberi kesan ke atas keterlibatan pelajar dalam e-pembelajaran di institusi pengajian.

Berdasarkan ulasan di atas, terdapat beberapa elemen antara muka yang boleh memberi kesan ke atas keterlibatan pelajar dalam e-pembelajaran. Namun begitu, terdapat elemen antara muka yang dikemukakan oleh penyelidik lepas berbeza antara satu sama lain. Sebagai contoh, kajian yang dijalankan oleh Adnan et al. (2015) menekankan elemen imej, bentuk, warna dan teks sebagai elemen yang mampu membuatkan pelajar terlibat secara konsisten dalam e-pembelajaran. Beliau menyatakan

MyJICT - Malaysian Journal of Information and Communication Technology Vol 1 2016, Issue 2 | e-ISSN 0127-7812 103

bahawa elemen warna merupakan elemen yang penting kerana ia mampu menarik minat pelajar untuk terus terlibat dalam e-pembelajaran dan elemen bentuk pula menjurus kepada penggunaan simbol bagi memudahkan pelajar memahami ikon atau logo tanpa perlu menggunakan teks yang banyak. Sebaliknya, Faghih et al. (2013) pula menekankan elemen imej, muzik dan animasi dalam antara muka e-pembelajaran bagi meningkatkan keterlibatan pelajar. Perbezaan elemen antara muka disebabkan oleh keperluan dan kehendak pengguna antara muka yang berbeza antara satu sama lain. Selain itu, latar belakang sosio budaya pelajar juga memainkan peranan dalam menentukan elemen antara muka e-pembelajaran. Ini kerana elemen yang tidak menepati kehendak pelajar menyebabkan mereka keberatan untuk terus terlibat dalam menggunakan e-pembelajaran.

Berdasarkan elemen antara muka yang dikemukakan oleh penyelidik lepas, elemen warna sering kali disebut oleh kebanyakan penyelidik seperti Jadual 2. Pemilihan warna yang sesuai adalah penting untuk memastikan kualiti ketepatan dan penyampaian maklumat dalam reka bentuk antara muka. Tambahan pula, pemilihan warna yang baik juga berkaitan dengan aspek psikologi manusia. Ini kerana warna berperanan mempengaruhi persepsi manusia (Ibrahim, Ashaari, & Wook, 2013). Penggunaan warna yang tidak sesuai boleh menyebabkan pengguna memberi pelbagai reaksi dan perasaan yang berbeza boleh memberi kesan terhadap keterlibatan pelajar. Kombinasi warna yang menarik pada antara muka boleh menimbulkan minat pelajar untuk menggunakan e-pembelajaran secara berterusan.

Selain itu, penyelidikan mengenai keterlibatan pelajar dalam komunikasi manusia dan komputer masih kurang. Keterlibatan pelajar sering kali dikaji dari aspek kehidupan harian pelajar seperti menghadiri kelas, aktif bertanya dan menyiapkan kerja sekolah tanpa melibatkan hubungan antara manusia dan komputer. Namun, definisi keterlibatan pelajar yang telah dibentangkan boleh dijadikan sebagai titik permulaan bagi mengimplementasikan keterlibatan pelajar dalam reka bentuk antara muka e-pembelajaran. Penyelidikan dalam menghasilkan reka bentuk antara muka yang menarik minat pelajar diharap dapat menyelesaikan masalah e-pembelajaran yang sering diabaikan tanpa digunakan semaksima mungkin.

9. KESIMPULAN

Reka bentuk antara muka e-pembelajaran adalah penting kerana ia berperanan dalam interaksi pelajar dengan e-pembelajaran. Faktor reka bentuk antara muka yang lemah adalah faktor utama yang menyebabkan pelajar kecewa untuk menggunakan e-pembelajaran dan seterusnya mengurangkan keterlibatan dan komitmen pelajar terhadap e-pembelajaran. Ramai penyelidik lepas mencadangkan agar memberikan lebih perhatian terhadap reka bentuk antara muka dalam menarik minat pelajar kerana reka bentuk antara muka sangat kritikal kerana ia berkait rapat dengan keberkesanan e-pembelajaran. Oleh itu, elemen antara muka memainkan peranan penting dalam membuatkan antara muka e-pembelajaran lebih menarik dan boleh digunakan. Dalam ulasan di atas, penulis telah

MyJICT - Malaysian Journal of Information and Communication Technology Vol 1 2016, Issue 2 | e-ISSN 0127-7812 104

menyenaraikan lima elemen antara muka e-pembelajaran yang mampu meningkatkan keterlibatan pelajar terhadap e-pembelajaran iaitu elemen warna, imej, tipografi, navigasi dan reka letak. Berdasarkan tinjauan literatur, lima elemen yang dipilih adalah yang paling kerap muncul dalam penyelidikan lepas namun, elemen tersebut mungkin akan berubah setelah kajian empirikal dijalankan kelak.

Tambahan pula, penyelidikan mengenai keterlibatan pelajar dalam persekitaran bagi interaksi antara manusia dan komputer masih lagi kurang dan tidak mendapat perhatian seperti mana yang dibincangkan di atas. Oleh itu, mengimplementasikan budaya dalam reka bentuk antara muka e-pembelajaran merupakan satu jalan bagi meningkatkan keterlibatan dan komitmen pelajar terhadap e-pembelajaran. Ini kerana kehendak dan keperluan pelajar yang terdiri daripada latar belakang yang berlainan terhadap reka bentuk antara muka adalah berbeza antara satu sama lain dan menimbangkan perbezaan ini ketika membangunkan antara muka e-pembelajaran adalah penting bagi memastikan reka bentuk antara muka tersebut mampu menyampaikan pembelajaran yang berkesan kepada para pelajar. Selain itu, mengabaikan peranan budaya dalam antara muka boleh menyebabkan reka bentuk antaramuka tidak menepati kehendak pelajar dan seterusnya menolak daripada menggunakan e-pembelajaran. Terdapat pelajar yang tidak menggunakan e-pembelajaran jika reka bentuk antara muka tidak menggambarkan budaya mereka. Oleh itu, kajian ini berbaloi untuk diterokai dengan lebih mendalam dalam meningkatkan keberkesanan dan kejituan antara muka e-pembelajaran dan seterusnya meningkatkan keterlibatan pelajar menggunakan e-pembelajaran.

RUJUKAN

Adnan, A. S., Ali, M., & Ahmad, R. (2015). The Utilisation of Visual Elements on Interface Design

of e-learning . In International Conference on Information Technology & Society (pp. 273–279). Ali, A., & Smith, D. (2015). Comparing Social Isolation Effects on Students Attrition in Online

Versus Face-to-Face Courses in Computer Literacy, 12, 11–20.

Alias, N., Zakariah, Z., Ismail, N. Z., & Aziz, M. N. A. (2012). E-Learning Successful Elements for Higher Learning Institution in Malaysia. Procedia - Social and Behavioral Sciences, 67(November 2011), 484–489. doi:10.1016/j.sbspro.2012.11.353

Altaboli, A., & Lin, Y. (2011). Investigating effects of screen layout elements on interface and screen

design aesthetics. Advances in Human-Computer Interaction, 2011. doi:10.1155/2011/659758 Angelino, L. M., Williams, F. K., & Natvig, D. (2007). Strategies to Engage Online Students and

Reduce Attrition Rates. The Journal of Educators Online, 4(2), 1–14.

Arockiam, L., & Selvaraj, J. C. (2013). User Interface Design for Effective E-Learning based on Personality Traits. International Journal of Computer Applications, 61(14), 28–32.

Backs, A. (2012). Dragging the Y Generation Into E-Learning : A Study of Engaging Averse Online

MyJICT - Malaysian Journal of Information and Communication Technology Vol 1 2016, Issue 2 | e-ISSN 0127-7812 105

Learners (pp. 30–40).

Bakar, Z. A., & Long, P. (2013). A Study of Visual Appeal Interfaces Based on Subjective Preferences, (November), 25–26.

Bart, M. (2012). Online Student Engagement Tools and Strategies.

Barzegar, S., Shojafar, M., & Keyvanpour, M. R. (2009). User interface in virtual learning environment. Internet Technology and Secured Transactions, 2009. ICITST 2009. International Conference for, 1–6. doi:10.1109/ICITST.2009.5402572

Benaida, M. (2014). Developing Arabic Usability Guidelines for E-Learning Websites in Higher Education.

Boton, E. C., Wales, N. S., Gregory, S., & Wales, N. S. (2015). Minimizing Attrition in Online Degree Courses. Journal of Educators Online, 62–90.

Chakrabarty, S., & Mohamed, N. (2013). Problem Based Learning: Cultural Diverse Students ’ Engagement , Learning and Contextualized Problem Solving in a Mathematics Class, 2013(Singaram 2010), 38–49.

Ching, G. (2012). Factors affecting student engagement: An analysis on how and why students learn, (December).

Delialioǧlu, Ö. (2012). Student engagement in blended learning environments with lecture-based and problem-based instructional approaches. Educational Technology and Society, 15(3), 310–322. doi:10.2307/jeductechsoci.15.3.310

Dixson, M. D. (2010). Creating effective student engagement in online courses: What do students find engaging? Journal of the Scholarship of Teaching and Learning, 10(2), 1–13.

Dunleavy, J., & Milton, P. (2009). What Did You Do in School Today? Exploring the Concept of Student Engagement and Its Implications for Teaching and Learning in Canada. Concept Paper May 2009.

Essam, S., & Al-Ammary, J. (2013). The Impact of Motivation and Social Interaction on the E

Learning at Arab Open University, Kingdom of Bahrain. Creative Education, 04(10), 21–28. doi:10.4236/ce.2013.410A004

Faghih, B., Azadehfar, M. R., & Katebi, S. D. (2013). User Interface Design for E-Learning Software. The International Journal of Soft Computing and Software Engineering, 3(3), 786–794. doi:10.7321/jscse.v3.n3.119

Fredricks, J., McColskey, W., Meli, J., Mordica, J., Montrosse, B., & Mooney, K. (2011). Measuring student engagement in upper elementary through high school: a description of 21 instruments. Issues and Answers Report, 098, 26–27. Retrieved from http://ies.ed.gov/ncee/edlabs

Galitz, W. O. (2007). The Essential Guide to An Introduction to GUI Design Principles and Techniques. Xtemp01.

Guralnick, D. (2006). User Interface Design for Effective, Engaging E-Learning (pp. 1–10). Retrieved

MyJICT - Malaysian Journal of Information and Communication Technology Vol 1 2016, Issue 2 | e-ISSN 0127-7812 106

from http://www.kaleidolearning.com/Effective E-Learning.pdf

Ibrahim, M. A., Ashaari, N. S. @, & Wook, T. S. M. T. (2013). Colour Impressions Framework. Procedia Technology, 11(Iceei), 305–311. doi:10.1016/j.protcy.2013.12.195

Kamaruddin, N. (2010). Challenges of Malaysian Developers in Creating Good Interfaces for

Interactive Courseware. The Turkish Online Journal of Educational Technology, 9(1), 37–42. Kamaruddin, N. (2012). Understanding Interface Design Role in Facilitating Teaching and Learning

Process: A Conceptual Framework, 1–11.

Kamaruddin, N. (2014). Understanding Interface Design Role in Facilitating Teaching and Learning Process: A Conceptual Framework, 1–11.

Kamaruddin, N., Park, J. Y., & Nam, Y. H. (2012). The Quality of Interface Design for Educational Courseware Development in Malaysian Educational Context, XXXIII(2), 81–87. doi:10.1007/s13398-014-0173-7.2

Katuk, N. (2013). Progressive assessment of student engagement with web-based guided learning.

Interactive Technology and Smart Education, 10(2), 116. doi:10.1108/ITSE-11-2012-0023

Khanum, M. A., Fatima, S., & Chaurasia, M. A. (2012). Arabic Interface Analysis Based on Cultural

Markers. International Journal of Computer Science Issues (IJCSI), 9(1), 255.

Kim, I., & Kuljis, J. (2010). Manifestations of Culture in Website Design, 125–132.

Ladbrook, J., & Parr, J. (2015). Motivation, Leadership and Curriculum design. In Motivation, Leadership and Curriculum design (pp. 161–172). doi:10.1007/978-981-287-230-2

Lim, Y. M., Ayesh, A., & Chee, K. N. (2013). Socio-Demographic Differences in the Perceptions OF Learning Management System. International Journal of Software Engineering and Applications, 4(5), 15–35. Retrieved from http://airccse.org/journal/ijsea/papers/4513ijsea02.pdf

Mcfarlane, P. S. (2011). Determining Student Preference for a Culturally Specific e-Learning Website.

Müller, R. (2014). Generation Y Students’ Product Colour Preferences. Mediterranean Journal of Social Sciences, 5(21), 69–74. doi:10.5901/mjss.2014.v5n21p69

Newmann, F. M., Wehlage, G. G., & Lamborn, S. D. (1992). Student Engagement and Achievement in American Secondary Schools. Student engagement and achievement in American secondary schools.

Parsons, J., & Taylor, L. (2011). Student Engagement: What do we know and what should we do ? The University Partners, (March), 1–59. Retrieved from https://education.alberta.ca/media/6459431/student_engagement_literature_review_2011.pdf

Ramakrisnan, P., Jaafar, A., Razak, F. H. A., & Ramba, D. A. (2012). Evaluation of user Interface Design for Leaning Management System (LMS): Investigating Student’s Eye Tracking Pattern and Experiences. Procedia - Social and Behavioral Sciences, 67, 527–537. doi:10.1016/j.sbspro.2012.11.357

MyJICT - Malaysian Journal of Information and Communication Technology Vol 1 2016, Issue 2 | e-ISSN 0127-7812 107

Reyna, J. (2009). Developing quality e-learning sites: A designer approach. In AScilite 2009 (pp. 837–

838). Retrieved from http://www.ascilite.org/conferences/auckland09/procs/reyna-1-poster.pdf Reyna, J. (2013). The importance of visual design and aesthetics in e-learning. Australian Institute of

Training and Development, (40), 28–32.

Shahzad, A., Hafizi, W., & Golamdin, A. G. (2014). E-Learning Navigation Model Based on Student ’ s Learning Behavior: Case Study in UUM. International Journal of Business and Social Science, 5(5), 118–130.

Sithole, K., B. D., I., & Onyari, E. K. (2012). Influence of Generations’ Traits on teaching and Learning in an Open Distant Learning (ODL) Environment, 1–9.

Thomson, S. (2014). 9 Tips To Improve Online Learners’ Engagement. Retrieved from https://elearningindustry.com/9-tips-improve-online-learners-engagement

Tomczyk, L. (2009). Graphical User Interface in e-Learning System Dedicated for Seniors, 617–622. doi:10.1007/978-0-387-89024-1

Trowler, V. (2010). Student engagement literature review. Higher Education, (November), 1–15. Retrieved from http://americandemocracy.illinoisstate.edu/documents/democratic-engagement-white-paper-2_13_09.pdf

Wang, C., Ke, S., Chuang, H., Tseng, H., & Chen, G. (2010). E-learning system design with humor and empathy interaction by virtual human to improve students’ learning. In Proceedings of the 18th International Conference on Computers in Education. Putrajaya, Malaysia: Asia-Pacific Society for Computers in Education (pp. 615–622).

Welch, B. K., & Bonnan-White, J. (2012). Twittering to increase student engagement in the university classroom. Knowledge Management and E-Learning, 4(3), 325–345.

Yaa, N., & Adzobu, A. (2014). Design, Use and Evaluation of E-Learning Platforms: Experiences and Perspectives of a Practitioner from the Developing World Studying in the Developed World, 147–159. doi:10.3390/informatics1020147

Zaharias, P. (2008). Cross-cultural differences in perceptions of E-learning usability: An empirical

investigation. International Journal of Technology and Human Interaction, 4(3), 1–26.

Retrieved from http://www.scopus.com/inward/record.url?eid=2-s2.0-56249091448&partnerID=40&md5=2b4dab9138916809f8018a192115af9b

Zhang, Y. (LeaF). (2013). Power Distance in Online Learning: Experience of Chinese Learners in U . S . Higher Education. The National Review of Research in Open Distance Learning, 14(4), 239 – 254.

MyJICT - Malaysian Journal of Information and Communication Technology Vol 1 2016, Issue 2 | e-ISSN 0127-7812 108

An applicant for a driver license (DL) or identification card (ID) must present proof of lawful presence in the US. The table on the following pages describes the acceptable documents for each type of applicant attempting to verify lawful presence. All documentation must show the applicant’s name and date of birth. The applicant must validate a name change or other inconsistent information through additional documentation such as a marriage license, divorce decree or court order.

The department must verify applicable lawful presence documentation through the US Department of Homeland Security’s (DHS) Systematic Alien Verification for Entitlements (SAVE) Program. Verification through SAVE is often instantaneous, but when it is not, receipt of the DL/ID may be delayed for up to 30 days. If SAVE cannot verify on the first attempt, SAVE will permit two additional stages of verification. Each stage may require additional documentation from the applicant. After each stage, the applicant will receive instructions either verbally or by mail on how to proceed with the transaction. To avoid further delay, the applicant should comply with the instructions fully and as soon as possible. If the applicant provides timely responses, the process timeline generally occurs as follows.

Stage DLD receives response

from DHS DLD response to applicant

First Within a few seconds If verified, card issued

Second 3 to 5 business days Instruction letter issued within 48 hours after DHS response received by DLD

Third Up to 20 additional business days after second response received from DHS Instruction letter issued within 48 hours after DHS response received by DLD

Temporary Visitor/Limited Term Issuance

An applicant may be issued a limited term DL/ID if he or she is NOT:

A US citizen;

A US national;

A lawful permanent resident;

A refugee; or

An asylee.

A limited term DL/ID will expire with the applicant’s lawful presence as determined by DHS. Commercial Driver Licenses

This guide does not apply to commercial driver licenses. A person who is a US citizen, US national, lawful permanent resident, refugee or asylee may apply for a commercial driver license. All others may apply for a nonresident commercial driver license, if eligible. Refer to http://www.dps.texas.gov/DriverLicense/CommercialLicense.htm or Chapter 522 of the Transportation Code for application and eligibility requirements.

(Rev. 7-13)

Category Acceptable Documents

U.S. Citizen v Birth certificate issued by the appropriate vital statistics agency of a U.S. State, a U.S. territory, or the District of Columbia indicating birth in U.S.

v Department of State Certification of Birth issued to U.S. Citizens born abroad (FS-240, DS-1350, or FS-545) or Consular Report of Birth Abroad

v Certificate of U.S. Citizenship

v Certificate of Naturalization

v U.S. Dept. of Justice – INS U.S. Citizenship Identification Card (I-197 or I¬179)

v Northern Mariana Card (I-873)

v U.S. passport book that does not indicate on the last page that "THE

BEARER IS A UNITED STATES NATIONAL AND NOT A UNITED STATES CITIZEN"

v U.S. passport card

U.S. National U.S. passport book that indicates on the last page that "THE BEARER IS A UNITED STATES NATIONAL AND NOT A UNITED STATES CITIZEN"

Kickapoo Traditional Tribe of Texas (“KIC”) (U.S. citizen) American Indian Card (form I-872) which indicates “KIC”

Kickapoo Traditional Tribe of Texas (“KIP”) (non-U.S. citizen) American Indian Card (form I-872) which indicates “KIP”

American Indian born in Canada (First Nations) An applicant may refer to the Jay Treaty, 8 U.S.C. § 1359, or 8 C.F.R. § 289.2 and may present a variety of documents. Issuance cannot occur without approval of the documents by Austin headquarters. DLD Personnel: make copies of documentation and seek approval through the chain of command.

Lawful Permanent Resident v Permanent Resident Card (I-551)

v Resident Alien Card (I-551) – card issued without expiration date

v Valid Immigrant Visa (with adit stamp) and unexpired foreign passport

v Unexpired foreign passport stamped with temporary I-551 language (adit stamp), "Approved I-551," or "Processed for I-551"

v I-94 stamped with temporary I-551 language (adit stamp), "Approved I¬551," or "Processed for I-551"

v Re-entry Permit I-327

Note: I-151, the predecessor to I-551, is not acceptable as proof of permanent resident status.

Immigrant Visa with Temporary I- 551 language A valid Immigrant Visa within one year of endorsement (i.e. stamped by Customs and Border Protection – adit stamp) and an unexpired passport

Conditional entrants Immigration documentation with an alien number or I-94 number indicating this status, which can include but is not limited to:

v I-94 or other document showing admission under Section 203(a)(7), “refugee conditional entry”

v I-688B coded 274a.12(a)(3)

v I-766 with category A3 or A03

Category Acceptable Documents

Asylee Immigration documentation with an alien number or I-94 number indicating this status, which can include but is not limited to:

v I-94 with annotation “Section 208” or “asylee”

v Unexpired foreign passport with annotation “Section 208” or “asylee”

v I-571 Refugee Travel Document

v I-688B coded 274a.12(a)(5)

v I-766 with category A5 or A05

Refugee Immigration documentation with an alien number or I-94 number indicating this status, which can include but is not limited to:

v I-94 with annotation “Section 207” or “refugee”

v Unexpired foreign passport with annotation “Section 207” or “refugee”

v I-571 Refugee Travel Document

v I-688B coded 274a.12(a)(3)

v I-766 with category A3 or A03

Temporary Protected Status (TPS) Immigration documentation with an alien number or I-94 number indicating this status or Employment Authorization Document (EAD) (I-766) with category A12 or C19

Applicant with Employment Authorization Document Employment Authorization Document (EAD)( I-766)

Applicants for adjustment of status

Note: These are individuals applying to become lawful permanent residents. Immigration documentation with an alien number or I-94 number

This can include but is not limited to a form I-797 indicating pending I-485 or pending application for adjustment of status.

Applicants for extension of status, change of status, petition for non-immigrant worker, with a pending I-918 application, or other pending category. Immigration documentation with an alien number or I-94 number

This can include but is not limited to a form I-797 indicating a pending application for an extension of status, change of status, petition for non-immigrant worker, or other pending category.

Citizens of the Republic of Palau Unexpired foreign passport or I -94 with annotation “CFA/PAL” or other annotation indicating the Compact of Free Association/Palau

Employment Authorization Document (EAD)(I-766) with category A8 or A08

Citizens of the Republic of the Marshall Islands Unexpired foreign passport or I -94 with annotation “CFA/RMI” or other annotation indicating the Compact of Free Association/Republic of Marshall Islands

Employment Authorization Document (EAD)(I-766) with category A8 or A08

Category Acceptable Documents

Citizens of the Federated States of Micronesia Unexpired foreign passport or I -94 with annotation “CFA/FSM” or other annotation indicating the Compact of Free Association/Federated States of Micronesia

Employment Authorization Document (EAD)(I-766) with category A8 or A08

Cuban/Haitian entrants Immigration documentation with an alien number or I-94 number

This can include but is not limited to an I-94 with annotation “Cuban/Haitian entrant”

Lawful temporary residents Immigration documentation with an alien number or I-94 number

Self-petitioning abused spouses or children, parents of abused children, or children of abused spouses

(Applicants with Violence Against Women Act (VAWA) petitions) Immigration documentation with an alien number or I-94 number

This can include but is not limited to I-797 indicating approved, pending, or prima facie determination of I-360 or an approved or pending I-360 or an I-766 with category C31.

Parolees Immigration documentation with an alien number or I-94 number

This can include but is not limited to an I-94 with annotation “parole” or “paroled pursuant to Section 212(d)(5).”

Person granted deferred action Immigration documentation with an alien number or I-94 number

Persons granted deferred enforcement departure (DED) Immigration documentation with an alien number or I-94 number or Employment Authorization Document (EAD) (I-766) with category A11

Note: Individuals in this status may have been granted an extension to the

period of authorized stay that is not reflected on the current EAD. Notifications regarding any extensions to this category will be distributed by Austin headquarters.

Person granted family unity Immigration documentation with an alien number or I-94 number

Persons under an order of supervision Immigration documentation with an alien number or I-94 number

Persons granted extended or voluntary departure Immigration documentation with an alien number or I-94 number

Category Acceptable Documents

Persons granted withholding of deportation or removal Immigration documentation with an alien number or I-94 number

This can include but is not limited to an I-94 or passport with annotation “Section 243(h)” or a letter or order from USCIS or court granting withholding of deportation or removal.

Persons in removal or deportation proceedings Immigration documentation with an alien number or I-94 number

Persons granted a stay of deportation Immigration documentation with an alien number or I-94 number

Persons granted voluntary departure Immigration documentation with an alien number or I-94 number

A-1, A-2, and A-3 Unexpired foreign passport or I -94

Note: Issuance cannot occur unless applicant presents a letter from U.S. Department of State with original signature indicating ineligibility for

Department of State issued driver license or requesting issuance of a state

issued identification card.

B1/B2 Visa/BCC with I-94

(Border Crosser Card , DSP-150, or “laser visa”) All of the following:

¨ Unexpired foreign passport,

¨ Visa (border crosser card), and

¨ I -94

Note: Applicant must have an I-94 to be eligible because of the time and

distance from the border restrictions for applicants who do not obtain an I-94.

B-1, B-2, C-1, C-3, D-1, and D-2 Unexpired foreign passport or I -94

Note: The applicant may not be able meet residency/domicile requirements.

C-2

Alien in transit to U.N. Headquarters district. Travel limited to 25 miles radius of Columbus Circle in New York, NY This status is restricted to New York, NY and not eligible for a Texas driver license under the domicile/residency requirements.

E-1, E-2, and E-3 Unexpired foreign passport or I -94

E-2 CNMI

Treaty-investor and dependents in Commonwealth of the Northern Mariana Islands This status is limited to persons entering the Commonwealth of the Northern Mariana Islands (CNMI) and is not eligible for a Texas driver license (8 CFR § 214.2(3)(23)).

F-1

Foreign academic student Unexpired foreign passport or I -94 or I-20

Category Acceptable Documents

F-2

Dependent on F-1 Unexpired foreign passport or I -94

F-3

Commuter Student from Canada or Mexico This status is for commuters from Mexico or Canada and is not eligible for a Texas driver license under the domicile/residency requirements.

G-1, G-2, G-3, G-4, and G-5 Unexpired foreign passport or I -94

Note: Issuance cannot occur unless applicant presents a letter from US Department of State approving the issuance of a DL/ID.

H-113, H-1131, H-1C, H-2A, H- 213, H-2R, H-3, H-4, and I Unexpired foreign passport or I -94

]-1

Exchange visitor (may be student, trainee, work/travel, au pair, etc.) Unexpired foreign passport or I -94 or DS-2019

]-2

Dependent of J-1 exchange visitor Unexpired foreign passport or I -94

K-1, K-2, K-3, K-4, L-1, L-1A, L- 113, and L-2, Unexpired foreign passport or I -94

M-1

Non-academic student Unexpired foreign passport or I -94 or I-20

M-2

Dependents of non-academic students Unexpired foreign passport or I -94

M-3

Commuter Student from Canada or Mexico This status is for commuters from Mexico or Canada and is not eligible for a Texas driver license under the domicile/residency requirements.

N-1 through N-7 (NATO)

North American Treaty

Organization

Representatives and dependents Unexpired foreign passport or I -94

N-8, N-9, O-1, O-2, O-3, P-1, P

2, P-3, P-4, Q-1, Q-2, Q-3, R-1, R-2, S-5, S-6, S-7, T-1, T-2, T-

3, T-4, T-5, TN-1, TN-2, TD, U-1, U-2, U-3, U-4, U-5, V-1, V-2, and V-3 Unexpired foreign passport or I -94

Category Acceptable Documents

WB*

Visitor for business (visa waiver program) Unexpired foreign passport with admission stamp annotated “WT/WB” or I -94 Note: The applicant may not be able meet residency/domicile requirements.

WT*

Visitor for pleasure (tourist in visa waiver program) Unexpired foreign passport with admission stamp annotated “WT/WB” or I -94 Note: The applicant may not be able meet residency/domicile requirements.

*Visa waiver program countries: Andorra, Australia, Austria, Belgium, Brunei, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Japan, Latvia, Liechtenstein, Lithuania, Luxembourg, Malta, Monaco, the Netherlands, New Zealand, Norway, Portugal, San Marino, Singapore, Slovakia, Slovenia, South Korea, Spain, Sweden, Switzerland, and the United Kingdom.

European Journal of Business and Management www.iiste.org

ISSN 2222-1905 (Paper) ISSN 2222-2839 (Online) Vol.5, No.22, 2013

Challenges Facing the Development of Islamic Banking. Lessons

from the Kenyan Experience

Shamim Njeri Kinyanjui

Jomo Kenyatta University of Agriculture and Technology

Abstract

The purpose of the study was to investigate the challenges facing the development of Islamic banking in Kenya. The research designed for study was a case study approach focusing on four Islamic compliant banks in Kenya. These are: - First Community Bank; Gulf African Bank; Dubai Bank; Kenya commercial Bank Ltd and Barclays Bank Ltd. The population of the study consisted of 33 customers, who were the holders of accounts in the respective banks and 11 managers. Sample data collected by use of questionnaires administered by the researcher and a research assistant. Data analysis method used is based on the quantitative approach using descriptive statistics: mean, mode, and median. Frequency tabulations and cross tabulations were used to bring out the finding of the study. The study revealed Islamic banking compliant was driven by religious compliance and customers need being met. It also revealed that continuous review and improvement of shariah compliant products together with diversifying market niche will lead to drastic development and marketing of Islamic banking products. From the study the following conclusions were drawn: firstly, the factors that influence development of Islamic banking products in Kenya are purely religious compliance and customers need being met.

Keyword: Islamic Banking.

1.1 Introduction

There is no satisfactory definition of a bank and term bank, as such, may apply to an organization, a financial institution operated by an individual as sole proprietor, partnership, a corporation or other type of association, in modern use it is an establishment for the custody of money (Saddiqi, 1986).

One definition of an Islamic Bank is a bank that, by its own choice, opts to comply with two sets of law: the law of the Land (Jurisdiction); and the Islamic Law (Shari'ah). This is why Islamic bankers have two types of legal counsel: traditional ̏lawyers" and ̏Shari'ah Councils" (Al-Bahar, 1996). As Islamic finance is intertwined within its religion, the basis of the religion affects the finance in two important ways: ̏Islam aims at building a socio-economic order based on justice and considers economic activity as a means to an end and not an end in itself. It enjoins Muslims to harness natural resources, which are a trust from Allah( God) , for carrying out rightful activities; but abhors exploitation and man-made inequalities of income and wealth." ̏Islam is deeply concerned with the problem of economic development, but treats this as an important part of a wider problem, that of total human development. The primary function of Islam is to guide human development on correct lines and in the right direction. It deals with all aspects of economic development but always in the framework of total human development and never in a form divorced from this perspective" (Al-Harran, 1993).

Islamic banking has the same purpose as conventional banking except that it operates in accordance with the rules of Shari’ah, known as Fiqh al-Muamalat (Islamic rules on transactions). The basic principle of Islamic banking is the sharing of profit and loss and the prohibition of riba (usury/interest). Common terms used in Islamic banking include; profit sharing (Mudharabah), safekeeping (Wadiah), joint venture (Musharakah), cost plus (Murabahah), and leasing (Ijarah). In an Islamic mortgage transaction, instead of loaning the buyer money to purchase the item, a bank might buy the item itself from the seller, and re-sell it to the buyer at a profit, while allowing the buyer to pay the bank in installments. However, the bank's profit cannot be made explicit and therefore there are no additional penalties for late payment. In order to protect itself against default, the bank asks for strict collateral. The goods or land is registered to the name of the buyer from the start of the transaction. This arrangement is called Murabaha. Another approach is EIjara wa EIqtina, which is similar to real estate leasing. Islamic banks handle loans for vehicles in a similar way by selling the vehicle at a higher-than-market price to the debtor and then retaining ownership of the vehicle until the loan is paid (Kahf, 2004).

An interest free Islamic bank in relation to its clients plays the role of partner, investor and trader. The interest free modes of operation designed by Islamic financial institutions satisfy the criteria „God permits trading and forbid riba. Keeping in view of the Islamic injunctions against taking and giving of interest, the Islamic banks

European Journal of Business and Management www.iiste.org

ISSN 2222-1905 (Paper) ISSN 2222-2839 (Online) Vol.5, No.22, 2013

have organized their operations on the basis of profit /loss sharing which is permitted in Islam. Islamic banks accept demand deposits and time deposits. Demand deposit are fully repayable on demand and do not get any return, holder of time deposits are given a share in the profit earned by the bank according to a profit sharing rates made known in advance.

1.2 Problem Statement

The study carried out by Gerrald in the year 1997 found that: Islamic religious belief and social responsibility are the two most important factors that determine bank selection. Cost benefit is the third most important factor considered in bank selection. Clients of conventional and Islamic banks share a number of motives, but they differ significantly on a few motives in relation to bank selection, the clients of Islamic banks are more familiar with the products/services that conform to the Shari’ah. However, the study by Erol and El-Bdour (1989) contradict the findings of other studies and argue that religious motivation did not appear to play a primary role in bank selection. Significant number of customers would withdraw their deposits if an Islamic bank did not generate sufficient profit to make a distribution in any one year hence motivated by higher dividend payments instead. Further, the results of the study by Oundo (2009) suggested that there was poor supply of Shari’ah-compliant products in Kenya’s financial institution. The challenges in the supply side of Shari’ah compliant financial services was illustrated by one bank that wrote Cheques to suppliers for their Muslim clients who were uncomfortable with receiving credit in cash against the Shari‟ah. There was high demand for Islamic financial products against a negligible supply of the same meaning that Islamic finance clients had few options (if any) and financially neglected by the microfinance sector. According to Ndungu (2010) Kenya was the first country in the East and Central African region to introduce Islamic banking in the year 2007. In this short period, two banks were licensed to exclusively offer Shari‟ah-compliant products with many other conventional banks establishing a window specifically for Shari‟ah-compliant products. He noted that the concept of shari‟ah complaint banking has emerged as an alternative vehicle for mobilization and supply of finance. For example, the two banks have already contributed in development agenda of the country by participating in Shari‟ah-compliant (Sukuks) components of infrastructure bonds issued by the Central Bank of Kenya on behalf of the Government of Kenya and “Structured Sukuk” is expected to cover the bonds and T-Bills market in future. However, although the concept of Islamic Finance has generated a lot of interest and overwhelming support from both Muslim and non-Muslim population in Kenya, as a regulator, CBK has faced by certain challenges which need to be addressed.

In their short period of existence, Islamic banking in Kenya has shown very commendable performance commanding combined market share of the banking sector in terms of gross assets of 0.8%. Currently there are two Islamic banks operating in Kenya: Gulf African and First Community bank, which had a loan portfolio of 4.9-billion shillings, deposits totaling 7.5billion shillings and 27270 deposit accounts. These indicators point to the tremendous potential of this market niche, which has been previously untapped, largely comprising Muslims estimated to make up at least 15% of Kenya's population of 36-million (Muriri, 2009). To address the above challenges highlighted by Central Bank of Kenya, this study seeks to establish factors influencing the development of Islamic banking in Kenya.

2.0 Literature Review

2.1 Islamic Banking Instruments

The Islamic banking model primarily relies on the instruments of mudarabah (joint venture) and musharakah (equity participation) to eliminate interest from the financial sector and economy. Other interest-free instruments such as murabaha (deferred payment sale), ijarah (leasing), bai Salam (advance payment) and bai istisna (procurement engagement) are also used to enhance the practical scope, diversification and risk management capabilities of the Islamic banking system.

2.1.1 Profit and loss sharing (PLS) Theory

Islamic scholars treat PLS instruments, mudarabah and musharakah as a central pillar of the Islamic banking model. In mudarabah banking, the Islamic bank accepts funds from depositors under risk-sharing arrangements. The Islamic bank either directly invests these funds in profitable investments or extends them to entrepreneurs on a risk-sharing basis. The Islamic bank shares the profit or loss made on mudarabah ventures with its depositors. In musharakah banking, the Islamic bank contributes the depositors' funds to a joint enterprise with

European Journal of Business and Management www.iiste.org

ISSN 2222-1905 (Paper) ISSN 2222-2839 (Online) Vol.5, No.22, 2013

the client (an entrepreneur). Generally, the Islamic bank allows the client to manage all the affairs of a Musharakah business.

The Islamic bank and the client mutually share the profit or loss made on the Musharakah investment. In a typical PLS arrangement, an Islamic bank provides the risk capital to a firm in which professional managers are responsible for making strategic and operational decisions. The bank shares in profits and is liable to any financial loss. There is no serious problem with this arrangement if the bank is able, and is allowed, to monitor business operations of the firm. However, proper monitoring mechanisms are yet to be devised for PLS, especially in case of Mudaraba that does not provide any control rights to the financier (the Islamic bank in this case). Fiqh literature on this issue is quite out-of-date and needs serious reconsideration. For example, Saleh (1986) lists three rights and one responsibility of the financier in a Mudaraba arrangement. The rights include ensuring that the borrowing entrepreneur (firm) complies with the terms of the contract, sharing profits, and limited liability in case of loss. The sole responsibility is handing over the Mudaraba capital. He also outlines two rights and two responsibilities of the borrower.

The rights include conducting the business with an appropriate degree of freedom, and accounting decisions. The responsibilities are compliance with the terms of the contract, and liquidation of the Mudaraba business at the end of the contract. The modern use of Mudaraba as a mode of financing obviously requires more than such preliminary specification of rights and responsibilities. There is a need for construction of standardized PLS contracts, or bylaws, in the light of the legal frameworks of Muslim countries. A prominent feature of these bylaws should be definition of the rights and obligations of various officers or groups within the organizational structure. Similar bylaws should delineate the clauses related to performance of the borrowing firm compared with other firms in the same sector and, possibly, other firms.

2.1.2 Murabaha (deferred payment sale): theory and practice

Under the Murabaha arrangement, the client makes a promise to buy specified goods from the Islamic bank on a deferred payment basis. The Islamic bank purchases goods from the original supplier and sells them on to the client after adding its own profit margin. The legality of murabaha could not be established from the primary sources of Islamic Shari’ah, i.e. the holy Quran and Sunnah. The early Islamic jurists, such as Imam Malik (796) and Imam Shafi (820), approved murabaha sales but they did not refer to the increase in price in the case of deferred payment. Subsequently, certain other Islamic jurists, such as Sarakhsi (1091), Marghinani (1197) and Nawawi (1277), allowed the seller to charge a higher price in the deferred payment sale by characterizing it as a normal trade practice (Saadullah, 1994; Vogel and Hayes, 1998). Contemporary Islamic scholars have mixed opinions about the murabaha banking system. The majority of them have strong reservations about it because of its close resemblance to conventional banking practice.

2.1.3 Ijarah (lease financing): theory and practice

The features of Ijarah financing are very similar to those of conventional lease financing. However, unlike in the conventional lease contract, Shari’ah holds the leaser responsible for all damage, repairs, insurance and depreciation of the leased asset. The leaser should also bear the risk of uncertainty attached to the useful life of the leased asset. Islamic financial institutions mostly rely on leasing, known as Ijarah wa iqtina, for meeting financing needs in the real estate, retail, industry and manufacturing sectors. Leasing enjoys strong support from Shari’ah scholars and bears a close resemblance to conventional leasing (Iqbal, 2000).

2.1.4 Bai Salam (advance payment) and bai istisna (procurement engagement)

Bai salam and bai istisna are forward sale contracts in which the seller pays in advance the price of goods that are to be delivered to him at a specified future date. Bai salam was widely practiced in the Arabian agricultural sector long before the dawn of Islam. These instruments are best suited to meet the financing needs of the farming and manufacturing industries in the Islamic economy. Shari’ah stipulates that the terms and conditions of bai salam and bai istisna contracts, such as price, quantity and quality of goods, should be clearly determined without involving any features of interest, gharar (speculation) or dubious sale (Iqbal, 2000).

2.2 Challenges facing Islamic Banking in Kenya

Islamic banking industry has been trying for the last over two decades to extend its outreach to bring it at least to the level of conventional banking. But the absence of Shariah-compliant legal framework — needed to make interest-free banking acceptable (and create sound financial institutions) — is the major snag behind its low penetration in the financial market.

European Journal of Business and Management www.iiste.org

ISSN 2222-1905 (Paper) ISSN 2222-2839 (Online) Vol.5, No.22, 2013

It is the time to take stock of challenges faced by the Islamic banks as they need a number of supporting institutions/arrangements to perform functions which are being carried out by various financial institutions in the conventional framework. Attempts should be made to modify the existing structure to provide better products and quality service within the ambit of Islamic laws.

Some of the most important challenges facing the Islamic banking industry are identified as follows.

Legal Support: Islamic law offers its own framework for execution of commercial and financial contracts and transactions. Nevertheless, commercial banking and company laws appropriate for implementation of Islamic banking and financial contracts do not exist. Islamic banking contracts are treated as buying and selling properties and hence are taxed twice. The commercial, banking and company laws contain provisions that are narrowly defined and prohibit the scope of Islamic banking activities within conventional limits. It is necessary that special laws for the introduction and practice of Islamic banking be put in place. The legal framework of Islamic banking and finances might include the following:

a. Islamic banking courts: The disputed cases of the Islamic banks are subject to the same legal system and are dealt with the same court and judge as the conventional one while the nature of the legal system of Islam is totally different. To ensure a proper, speedy and supporting Islamic legal system, amendments in existing laws, which are repugnant to injunctions of Islam, are required to promulgate Shariah compliant law for resolution of disputes through special courts.

b. Amendment of existing laws: Islamic banking has some kind of resemblance to universal banking, therefore, laws and regulations have to be amended accordingly to accommodate this new concept such as sections 7 (forms of business in which the banking company can engage) and 9 (prohibition of trade) of the Banking Companies Ordinance 1962 while Islamic banks are big or wholesale traders in reality.

c. Islamic banking law: In the absence of Islamic banking laws, the enforcement of agreements in courts may require extra efforts and costs. Therefore, banking and companies’ laws in several countries require suitable modifications to provide a level playing field for Islamic banks. Furthermore, international acceptance of Islamic financial contracts requires them to be Shariah compatible as well as acceptable under the major legal regimes such as Common law and Civil law systems.

d. Islamic banking balance sheet: Islamic banks do not show assets financed through Ijara, Murabah etc., on balance sheet because section 7 of Banking Ordinance 1962 does not allow a bank to own property or asset which section 9 prohibits to enter into any kind of trade. However, all the assets owned by Islamic banks be mentioned in their balance sheets.

e. Monthly payment agreement: The housing finance is executed on the basis of Diminishing Musharaka by the Islamic banks. Under this mode the house is jointly owned by the bank and the customer. The bank rents out its share to the customer on Ijara basis. The Islamic bank while executing Ijara with the partner/customer uses the term ‘Monthly Payment Agreement’ instead of having the Ijara agreement with the customer.

It is so named as to safeguard the bank’s interest in case of refusal by the customer to pay rentals. No legal cover is provided to the Islamic bank to overcome this risk.

f. PLS deposits: Deposits in Islamic banks are usually based on principle of profit and loss (Musharaka or Murabaha). If something happens and the bank suffers loss it has to be transferred to the depositor directly.

This fear of loss is the biggest barrier to deposit mobilisation in Islamic banks. In some cases, it leads to withdrawal of funds. The depositors should be provided with some kind of protection.

Islamic prudential regulations: Supervision of Islamic banks is equally important. At present, lack of effective prudential regulation is one of the weaknesses of the Islamic banking industry. For instance, leasing prudential regulations are applied to Ijara where the nature of both is different, such as taking advances. The bank is the owner in Ijara; so taking advances will render the contract of Ijara for conversion into Musharakah whereas the rules of Ijara are applied to it, which is illegal. And some of the Islamic banks are using the term of security, hence making the Ijara contract non-Shariah compliant as using the deposited sum under the heading of Ijara security (‘Rahn’) is nothing but Riba which is strictly prohibited by Islam.

Moreover, Ijara financing is subject to compulsory insurance which is essentially prohibited.

European Journal of Business and Management www.iiste.org

ISSN 2222-1905 (Paper) ISSN 2222-2839 (Online) Vol.5, No.22, 2013

Risks: The nature of risk in Islamic banking is different from those of conventional banking and therefore some special prudential, accounting and auditing standards should be applied to them.

Benchmark: Taking the conventional interest based benchmarks (Kibor etc.,) as the base of pricing an Islamic financial product puts Islamic banks at the mercy of their conventional peers. A negative perception is created among the clientele that there is no prudent difference in Islamic bank products as these are also using the same interest based benchmark. The mechanism for long-term financing could be devised on the basis of prevailing renting system adopted by the private landlords while renting their assets/properties etc.

Shariah based product: All Islamic financial institutions offer the same basic products, (90 per cent Murabaha and Ijarah) but the problem is that each institution has its own group of Islamic scholars on the Shariah board to approve the product. Consequently, the very same product may have different features and will be subject to different kind of rules in these institutions.

Lack of standard financial contracts and products can be a cause of ambiguity and a source of dispute and cost. In addition, without a common understanding on certain basic foundations, further development of banking products is hindered.

Nature of Islamic banking: Islamic banks are offering only Murabaha and Ijarah while leaving the core and difference making Islamic financial instruments such as Musharakah and Murabah. It is necessary to enhance and facilitate the implementation of real Islamic banking activities i.e. promoting risk sharing through equity type facilities on the asset side and profit sharing investment accounts on the funding side.

Lender of last resort facility: Islamic banks are reluctant to enter into long-term transactions due the lack of availability of liquidation through secondary market. There is liquidity support in the form of lender of last resort facility.

There is no proper mechanism of transparency and disclosure to the public in order to ensure consumer protection as provided by Shariah.

Islamic future exchange: In conventional system, long-term finance is provided through long-term bonds and equities. Apart from the general public, the most important source of these long-term investments are investment banks, mutual funds, insurance companies and pension funds. Islamic banks do not deal with interest bearing bonds. Therefore, their need for equity markets is much higher. On the top of it, the most of the products in Islamic banks are based on goods and commodities while prices and currency rates go up and down frequently, creating a big risk for them being traders in reality especially in the case of Salam and Istisna'a. To hedge the risk, they are in need of derivative products and consequently of Future Exchanges.

Shortage of experts in Islamic banking: The supply of trained or experienced bankers has lagged behind the expansion of Islamic banking. The training needs affect not only domestic banks, both Islamic and non-Islamic, but foreign banks as well.

Absence of accounting (and auditing) standards pertinent to Islamic banks: Uncertainty in accounting principles involves revenue realization, disclosures of accounting information, accounting bases, valuation, revenue and expense matching, among others. Thus, the results of Islamic banking schemes may not be adequately defined, particularly profit and loss shares attributed to depositors.

Lack of uniform standards of credit analysis: Islamic banks have no appropriate standard of credit analysis. Similarly, there is a widespread training need involving related aspects such as financial feasibility studies, monitoring of ventures, and portfolio evaluation.

Potential conflicts with central banks: Islamic banks have been established as separate legal entities; therefore, their relationships with central banks and/or other commercial banks are uncertain. Problems may be further aggrivated when an Islamic bank is established in a non-Muslim nation like the case of Kenya, and is subject to that nation's rules and requirements.

Instruments that meet the demand of specific investment requirements: One of the biggest challenges facing institutions is the provision of short-term investment instruments. Several institutions have tried to develop high quality short-term instruments, but have been hampered by their ability to generate assets, by their credit ratings, and by liquidity.

European Journal of Business and Management www.iiste.org

ISSN 2222-1905 (Paper) ISSN 2222-2839 (Online) Vol.5, No.22, 2013

3.0 Methodology

3.1 Design

The research design employed in this study was descriptive in nature in form of a survey. According to Chadran, (2004) a descriptive design enables the researcher to collect in depth information about the population being studied. Study by Gerrald (1997) have 27 used a similar research design where five significant factors were noted, the most relating to non Muslim being the desire to be paid higher interest on their savings.

3.2 Sampling

The study employed a combination of census and probability sampling. The census method was used to include all managers in the banks whose customers were sampled. The probability sampling was used to sample the customers; it was on a walk in basis. It was assumed that since the bank customers walk in at random, interviewing all willing customers was random enough. According to Kothari (2004), a sample is a representative group of the accessible population that is selected using various sampling methods. The sample size for the managers was 11 and that of the customers was 33. This was gotten from all the Islamic Banks located in Mombasa Island.

3.3 Data Collection

This study employed use of primary data, and obtained from bank employees and bank customers, by use of questionnaire. The use of questionnaire was justified because they 28 provide an effective way of collecting information from a large literate sample in a short period of time and at a reduced cost than other methods. Moreover, questionnaires facilitate easier coding and analysis of data collected.

3.3 Findings

The findings found out that 87.9% of the respondents were Muslims and only 12.1% were non-Muslims as shown in figure 3.3.1below.

Figure 3.3.1 religious population

European Journal of Business and Management www.iiste.org

ISSN 2222-1905 (Paper) ISSN 2222-2839 (Online) Vol.5, No.22, 2013

To determine the factors that influenced the banks to start providing shariah compliant products and services, they were asked to rate the above factors on a five point likert scale where one represent no influence and five represent influence to a great extent. The findings shows that the managers were not sure about the influence from increased competition, Staff aggressiveness and commitment and having a conducive government regulations and support, all with a mean score of about three. The findings show that potential huge profit from Islamic banking and Service quality gap were the two major factors influencing entry. On the other hand corporate social responsibility (CSR), activity driven by public concern and high bank liquidity requiring investments, and diversification to cushion the risk of corporate borrowing does not have any influence. This is summarized in the table below.

Factors Mean Std. Dev

Increased competition in the traditional bank market 3.27 1.272

Potential Huge profit from Islamic banking 3.73 1.009

Service quality gap 3.91 0.831

High bank liquidity requiring investment 1.82 1.079

As CSR activity driven by public concern 2.36 0.924

Diversification to cushion the risk of public borrowing 2.18 0.982

Staff aggressiveness and commitment 2.64 1.206

Conducive government regulations and support 3.18 1.401

The statistics in the table below shows that the managers are not sure with laying down institutional policies and procedures to regulate Islamic products as a way of ensuring sustainability of Shari’ah compliant product/services for Commercial Banks. The managers also disagreed that separation of the funds between Islamic and Conventional funds could lead to sustainability of the Islamic compliant product or services. On the other hand, the managers agreed that the following were the factors that lead to sustainability of the products: increased number of customers; support from senior management and board; aligned Islamic banking products policy to the overall strategy of the bank; well versed shari’ah Board members; management and staff who understand 44

Islamic Banking Market dynamics; good customers’ services and establishment of accounting that comply with Islamic Shari’ah.

Factors Mean Std. Dev

Lowering Bank Charges 2.82 1.168

Increasing earnings on deposits and savings 2.09 1.136

Relaxing the collateral requirement during lending 2.91 1.044

Diversifying Islamic banking product 4.64 0.505

Continuous review and improvement of shariah compliant products 4.55 0.522

Bonus to marketing staff 2.73 1.555

Interest free lending 4.73 0.905

Profit loss sharing approach 4.64 0.505

5.0 Conclusion

Other than providing Islamic compliant products and services, factors such as: offering product and services that meet the customer’s needs, reduced borrowing cost, minimum requirements to open an account and service efficiency influences the development of Islamic banking. The Islamic banks should put these priorities at forefront and strengthening it. There should also be an amendment to the existing Banking Law. By doing so, they can also win a great number of customers from both Muslims and non-Muslims community. Above all, they should invest in branch expansion thus enabling the marketing Islamic banking products at the grassroots level.

Lastly, there are future market potentialities for the Islamic banking products. Many younger generations were seen going for these products and services. For this banking, system to develop and grow the banks should create awareness through seminars, workshops, and advertisement and prayer sermons quoting verses from Quran and sayings of the prophet (hadith) that concern Islamic Banking system.

100

European Journal of Business and Management www.iiste.org

ISSN 2222-1905 (Paper) ISSN 2222-2839 (Online) Vol.5, No.22, 2013

6.0 Recommendations

In Kenya, all banks offering Islamic banking have established their own separate Shari’ah Board to supervise and offer guidance to their respective banks on Islamic banking system. In principle, Shari’ah Boards have the authority to impose their viewpoint, but logistic considerations do not permit timely vetting and/or monitoring of all banking operations. In view of these, I recommend CBK to put in place a policy to establish a universal Shari’ah Boards to oversee Islamic banking operation in Kenya. Conventional banking system in Kenya are compelled to have a uniform financial reporting standard thus mixing their funds that are both conventional and Islamic compliant. In view of this, Muslim customers are discouraged to bank with such conventional banks, as it is not acceptable in terms of shari’ah. Therefore, I recommend CBK to allow such conventional bank to have separate financial reporting standards. Lack of qualified Islamic banking personnel is another major bottleneck in Kenya. Further training and Islamic finance education should be given to bank personnel to up skill them to offer quality service and appropriate advice to bank customers. I recommend banks to organize regular training and workshops by inviting well-versed Islamic scholars to educate bank personnel about Islamic banking and financing.

References

Alexakis, C.&,Tsikouras, A, (2009). Islamic finance regulatory framework : challenges lying ahead, International Journal of Islamic and Middle Eastern Finance and Management, Vol. 2 Iss: 2, pp.90 – 104

Ahmad, S.M. (1952) “Interest and Unemployment” Islamic Studies, VIII I, pp. 9-46.

Al-Harran, S.A.S. (1993). Islamic Finance: Partnership Financing,Pelanduk Publications, Malaysia, pp. 15-157. Amin, H. (2008). Determination of choice criteria in Malaysia’s Retail Banking. Kuala Terrengganu.Malaysia. Anwar, M. (1989). Modeling interest free economy; Genuine publication. West, New Dhephi.

Al-Arabi,(1966) “Contemporary Banking Transactions and Islam’s Views Thereon” Islamic Review, May, pp. 10-16.

Ariff, (1982) Monetary and Fiscal Economics of Islam, International Centre for Research in Islamic Economics, Jeddah.

Bashir,(2000) Assessing the Performance of Islamic Banks: Some Evidence from the Middle East Paper retrieved on June 12, 2006 from http://www.luc.edu.orgs/meea/volume3/revisedbashir.pdf.

Bryman, C. (2001).Improving Internal Control: A Practical Guide for Microfinance Institutions.

Brunei. Khan, M. M. & Bhatti,M. I., (2008) Development in Islamic banking: a financial risk-allocation approach.VOL.1 PP 40-51. Khan, M. M., & Bhatti ,M. I. (2006). Why interest-free banking and finance movement failed in Pakista.Humanomics Vol 22, 3 pp 145-161. Australia. Kothari, C.R. (2004). Research methodology: methods & techniques. New Delhi: New age international (P) limited publishers.

Chandran, E. (2004). Research methods: A quantitative approach with illustrations from Christian ministries. Daystar University, Nairobi.

Chapra, (1982) “Money and Banking in an Islamic Economy”. In Ariff, M. Ed. Monetary and Fiscal Economics of Islam, Jeddah: International Centre for Research in Islamic Economics.

Cooper, D., & Schindler, P. (2000). Business research methods. (8th edn) New Delhi, Tata Mc Graw Hill. Ebrahim,

C. R. Kothari (1990) Research Methodology, Methods and Techniques. New Delhi, Wishwa.

101

European Journal of Business and Management www.iiste.org

ISSN 2222-1905 (Paper) ISSN 2222-2839 (Online) Vol.5, No.22, 2013

Erol, C. & El-Bdour, R. (1989). “Attitudes, behaviour and patronage factors of bank customers towards Islamic banks”. International Journal of Bank Marketing, Vol. 7 No. 6, p. 31.

Gafoor (2007) Islamic Banking Paper retrieved on June 27, 2007 from .

http://users.bart.nl/~abdul/chap4.html.

Gerrald, P. (1997).Division of Banking and Finance, Anyang business school.Nyanyang Technology University. ”. International Journal of Bank Marketing, 15 /16 pp204-216.

Husain, I.( 2010).Keynote address delivered at the Conference on Islamic banking and Finance : Islam and the Future of Economics .Karachi .

Iqbal, M. ed. 2000. Islamic Banking and Finance: Current Developments in Theory and Practice.Islamic Foundation and IAIE.

Iqhal (1991) Stabilization and growth in an open Islamic economy,Review of Islamic economy vol 1(2): 1-20 Kahf, M. (2004).Success factor of Islamic Banks, Symposium on Islamic Banking and Finance.

Kumar,M., Kee, F. T., & Charles .V. (2010).International Journal of Quality & Reliability Management, Volume 27, Number 3, 2010, pp. 352-378(27).Emerald Group Publishing Limited. Mingala, M. S. (2002). Financial Markets and Institutions. Addison Wesley, New York.

Khan (1987) Theoretical studies in Islamic banking and Finance, Houston institute for research and Islamic studies.

Microfinance Network with GTZ, Technical Guide No. 1 CBK. (2010). Targets Gulf cash with compliance bonds: Business daily news. Nairobi. Kenya.

M.S,(2001).Islamic banking in Brunei Darussalam. University of Brunei Darussalam/National University of Singapore, Singapore, and Tan Kai Joo Kemuda.te for research and Islamic studies.

Mirakhor (1989) Monetary management in an Islamic Economy:International Monetary Fund Washington D.C

Nasim, (1998), ‘Key Concepts of Islamic Finance: Their Relevance to Consumer Finance Products’, Proceedings of the Second Harvard University Forum on Islamic Finance, Cambridge, Massachusetts: Harvard University, 291-93.

Nasim, Pervez (1999), ‘Various Islamic Modes of Home Financing’ in Islamic Banking and Finance, The Concept, the Practice and the Challenge (ed.), Imtiazuddin Ahmad, Indianapolis: Islamic Society of North America, 69-73.

Siddiqi, M.N. 1988 “Islamic Banking: Theory and Practice”. Islamic foundation.

102

This academic article was published by The International Institute for Science, Technology and Education (IISTE). The IISTE is a pioneer in the Open Access Publishing service based in the U.S. and Europe. The aim of the institute is Accelerating Global Knowledge Sharing.

More information about the publisher can be found in the IISTE’s homepage: http://www.iiste.org

CALL FOR JOURNAL PAPERS

The IISTE is currently hosting more than 30 peer-reviewed academic journals and collaborating with academic institutions around the world. There’s no deadline for submission. Prospective authors of IISTE journals can find the submission

instruction on the following page: http://www.iiste.org/journals/ The IISTE

editorial team promises to the review and publish all the qualified submissions in a fast manner. All the journals articles are available online to the readers all over the world without financial, legal, or technical barriers other than those inseparable from gaining access to the internet itself. Printed version of the journals is also available upon request of readers and authors.

MORE RESOURCES

Book publication information: http://www.iiste.org/book/

Recent conferences: http://www.iiste.org/conference/

IISTE Knowledge Sharing Partners

EBSCO, Index Copernicus, Ulrich's Periodicals Directory, JournalTOCS, PKP Open Archives Harvester, Bielefeld Academic Search Engine, Elektronische Zeitschriftenbibliothek EZB, Open J-Gate, OCLC WorldCat, Universe Digtial Library , NewJour, Google Scholar

Jeffrey Graves

Curriculum Vitae

EDUCATION

2014–Present Ph.D., Computer Science, Tennessee Tech University, Cookeville, TN. Adviser William Eberle

2011–2016 M.S., Mathematics, Tennessee Tech University, Cookeville, TN.

Thesis Walsh Functions as Group Characters

Adviser Rafal Ablamowicz

2007–2011 M.S., Computer Science, Tennessee Tech University, Cookeville, TN.

Thesis Source Code Plagiarism Detection Using a Graph-Based Approach

Adviser William Eberle

2003–2007 B.S., Computer Science, Tennessee Tech University, Cookeville, TN.

GPA 3.61/4.0

Honors Cum Laude

AWARDS AND HONORS

2014 Graduate Student Teaching Award, Mathematics Department, Ten¬nessee Tech University.

2013 Challenge Assistantship, Mathematics Department, Tennessee Tech University.

TEACHING EXPERIENCE

Computer Science

o Design and Analysis of Algorithms, Fall 2014, Spring 2015, Fall 2015

Mathematics

o College Algebra, Spring 2012, Fall 2012, Fall 2013

o Finite Math, Spring 2013, Spring 2014

o Pre-Calculus I, Fall 2013, Spring 2014

o Pre-Calculus II, Spring 2014

o Transitional Algebra, Fall 2013

RESEARCH EXPERIENCE

Summer 2016 Advanced Data and Workflows Intern, National Center for Compu¬tational Sciences, Oak Ridge Natural Laboratory.

Summer 2015 Computational Data Analytics Intern, Computational Sciences and Engineering Division, Oak Ridge Natural Laboratory.

2009–2010 Department of Energy Grant, Center for Manufacturing Research, Tennessee Tech University.

Development of an electrical capacitance tomography sensing system to detect defects that arise in lost-foam casting processes as well as the development of packet recovery techniques for use wireless sensor networks deployed in foundry environments.

2008–2011 Department of Homeland Security Grant, Computer Science De¬partment, Tennessee Tech University.

Focussed on applying graph-based anomaly detection methods to areas such as insider threat detection, intrusion detection, cargo shipping, and source code similarity metrics.

2007–2008 Vanderbilt eHealth Initiative Subgrant, Computer Science Depart¬ment, Tennessee Tech University.

Development of a framework to integrate electronic patient record systems. Implementation of a distributed patient information system for health care providers to enable the sharing of patient records between separate health care organization environments.

UNIVERSITY SERVICE

2015–2017 Graduate Executive Committee Member, College of Engineering, Tennessee Tech University.

2015–2017 Graduate Student Mentor, Computer Science Department, Tennessee Tech University.

RESEARCH COMMUNITY SERVICE

2015 SIAM Data Mining Conference (SIAM) External Reviewer.

ACADEMIC WORK EXPERIENCE

2014–Present Graduate Teaching Assistant, Computer Science Department, Ten¬nessee Tech University.

2013–2014 Adjunct Faculty, Mathematics Department, Tennessee Tech University.

2012–2013 Graduate Teaching Assistant, Mathematics Department, Tennessee Tech University.

2011–2012 Mathematics Tutor, Library Learning Commons, Tennessee Tech University.

NON-ACADEMIC WORK EXPERIENCE

2011–2014 Contract Programmer, Automate My Data, Algood, TN.

Wrote and maintained materials requirement planning software and billing systems. Work involved Java, Python, MySQL, Pervasive.SQL, HTML, CSS, and Java Servlets

TECHNICAL PROFICIENCIES

Proficient In

Languages C, C++, OpenMPI, OpenMP, Java, Python

Technologies BSD, Linux, MacOSX, LATEX

Familiar With

Languages C#, PHP, Lisp, SQL, XML, RDF/SPARQL, HTML, CSS, Java Script

Libraries OpenMPI, OpenMP

Technologies Apache Jena Fuseki, Stardog, MySQL, PostgreSQL, SQLite, Pervasive.SQL

RESEARCH INTERESTS

Computer Algorithms, Theory of Computation, Artificial Intelligence, Graph Based Science Anomaly Detection

Mathematics Graph Theory, Coding Theory, Abstract Algebra, Real Analysis Recreational Fractals, Strage Attractors

PUBLICATIONS

Journal Articles

William Eberle, Lawrence Holder, and Jeffrey Graves. Insider threat detection using a graph-based approach. Journal of Applied Security Research, 6(1):32–81, January 2011

Conference Papers

Sreenivas R. Sukumar, Larry W. Roberts, and Jeffrey A. Graves. A reasoning and hypothesis-generation framework based on scalable graph analytics. Proceedings of the Cray User Group Conference, 2016

Michael Okaro, Mohamed Abdelrahman, and Jeff Graves. Monitoring metal fill profile in lost foam casting process using capacitive sensors and metal fill time estimation. In Sensors Applications Symposium (SAS), 2011 IEEE, pages 76–81. IEEE, February 2011

William Eberle, Lawrence Holder, and Jeffrey Graves. Using a graph-based approach for discovering cybercrime. In Florida Artificial Intelligence Research Society Conference, May 2010

William Eberle, Lawrence Holder, and Jeffrey Graves. Detecting employee leaks using badge and network IP traffic. In IEEE Symposium on Visual Analytics Science and Technology, October 2009

Mike Rogers, Sheikh Ghafoor, and Jeff Graves. Braid: Distributed patient information for health care providers. In International Conference on Internet Computing, pages 132–138, July 2009

Workshop Papers

Ferrol Aderholdt, Jeffrey A. Graves, and Manjunath Gorentla Venkata. Parallelizing single source shortest path with OpenSHMEM. In OpenSH-MEM 2017: Fourth workshop on OpenSHMEM and Related Technologies, 2017

Rina Singh, Jeffrey A Graves, Sangkeun Lee, Sreenivas R Sukumar, and Mallikarjun Shankar. Enabling graph appliance for genome assembly. In Proceedings of the 2015 IEEE International Conference on Big Data, BIG DATA '15, pages 2583–2590, Washington, DC, USA, October 2015. IEEE Computer Society

Posters

Rina Singh, Jeffrey A. Graves, and Douglas A. Talbert. Complex patterns in dynamic attributed graphs. In Proceedings of the 25th International Conference Companion on World Wide Web, WWW '16 Companion, pages 105–106, Republic and Canton of Geneva, Switzerland, 2016. International World Wide Web Conferences Steering Committee

Mike Rogers, Sheikh Ghafoor, and Jeff Graves. A framework for inte¬grating electronic patient record systems. American Medical Informatics Association Annual Symposium, 2009

Preference-based evolutionary algorithm for airport

runway scheduling and ground movement

optimisation

Michal Weiszer

School of Engineering

University of Lincoln

Lincoln, United Kingdom

Email: mweiszer@lincoln.ac.uk

Jun Chen

School of Engineering

University of Lincoln

Lincoln, United Kingdom

Email: juchen@lincoln.ac.uk

Paul Stewart

Institute for Innovation

in Sustainable Engineering

Derby, United Kingdom

Email: p.stewart1@derby.ac.uk

Abstract—As airports all over the world are becoming more congested together with stricter environmental regulations put in place, research on optimisation of airport surface operations started to consider both time and fuel related objectives. However, as both time and fuel can have a monetary cost associated with them, this information can be utilised as preference during the optimisation to guide the search process to a region with the most cost efficient solutions. In this paper, we solve the integrated optimisation problem combining runway scheduling and ground movement problem by using a multi-objective evo¬lutionary framework. The proposed evolutionary algorithm is based on modified crowding distance and outranking relation which considers cost of delay and price of fuel. Moreover, the preferences are expressed in a such way, that they define a certain range in prices reflecting uncertainty. The preliminary results of computational experiments with data from a major airport show the efficiency of the proposed approach.

I. INTRODUCTION

With continuous growth of air traffic, which is predicted to carry double passengers in 2030 compared to 2013 [1], it is forecasted that without any action taken, many airports will become congested and air transportation as a whole will have a significant impact on the environment. As a result, research on management of aircraft movements on and in the proximity of airports attracted a lot of attention in general (e.g. [2], [3]) and airport surface operations in particular [4]. Specifically, the most research focused on optimisation of individual airport surface operations such as ground movement, runway schedul¬ing or gate assignment to maximise the utilisation of available resources.

The objective of runway scheduling problem is of¬ten expressed as a minimisation of delay, the number of changes compared to First-come-first-served (FCFS) sequence, makespan or their combination. A wide range of exact and heuristic methods has been employed to solve this problem including dynamic programming [5], hybrid tabu search [6], and genetic algorithms [7]. A detailed review of recent research on runway scheduling problem can be found in [8].

The objective of ground movement has been considered to be mainly the minimisation of the total taxi time or other time related objectives [4]. Approaches to solve the ground movement problem include integer programming [9], [10] or

a graph-based approach utilised in [11], [12]. In addition to the total taxi time, a few researchers started to consider also fuel consumption during ground movement as an objective of the multi-objective optimisation problem [13]. Since runway scheduling and ground movement are interconnected problems, recently both problems have been integrated together, min¬imising time related objective [14]–[16] or both time and fuel consumption [17] in a multi-objective manner.

As some of the above-mentioned approaches use multi-objective optimisation to minimise distinct objectives such as time and fuel at the same time, the result of such approach is a set of non-dominated solutions, from which the decision maker (DM) has to implement one solution. In practice, often is the case that there exist some preferences beforehand, such as the price of fuel or cost of a single minute of delay. However, as argued in [18], utilising this information to convert the multi-objective problem into a single-objective one is con-traproductive. By finding only a single solution, DM is unable to investigate the properties of optimal/near-optimal solutions respecting higher-level preference information. Instead, this preference can be utilised during the optimisation to guide the search to an preferred region of interest (RoI) on the Pareto front. Research on incorporating preferences into multi-objective optimisation, whether in a-posteriori manner when the preferences are known before the search or in an interactive approach when preference information is iteratively obtained from the DM during the search, has been very active in the last years. Preferences are often expressed as reference points, goal vectors or aspiration levels corresponding to desired levels of objective values. The dominance relation is modified according to the distance to the reference point in [19] or aspiration level satisfaction in [20]. An achievement scalarizing function taking into account reference point is used to prefer some solutions closer to RoI in [21]–[24]. The original optimisation problem is modified by a weight distribution function in [25]. A binary preference in terms of linguistic variables is used in [26]; the crowding distance in Nondominated Sorting Genetic Algorithm-II (NSGA-II) [27] is changed in order to incorporate reference direction (weights) [28] or reference point [18], [23].

However, the desired values of objectives expressed as a reference point or aspiration levels are often unknown to the DM before the optimisation. As the result, an initial run of the optimisation algorithm is necessary to discover the whole

Pareto front from which the DM can subsequently choose a RoI. On the other hand, if the preferences are modelled as weights, the extent of RoI is usually defined by some non-intuitive parameter during the search as in [28].

In this paper, we solve the integrated optimisation problem combining runway scheduling and ground movement problem, similar as in [17], by using a multi-objective evolutionary opti¬misation (EMO) framework based on genetic algorithm (GA) and NSGA-II. The proposed algorithm employs a modified crowding distance presented in [23] and outranking relation of the Light Beam Search (LBS) method [29], which considers preference information in terms of cost of delay and price of fuel, guiding the search towards the most cost optimal RoI. Moreover, the preferences are expressed in a such way, that they give the DM a possibility to define a certain range of prices reflecting uncertainty.

The rest of the paper is organised as follows. Section II provides details about individual components of the integrated model, including runway scheduling and ground movement problem. The proposed EMO framework taking into account preferences is described in Section III. A set of computational experiments is carried out using data instances from Doha International Airport in Section IV. Finally, conclusions are drawn in Section V.

II. PROBLEM DESCRIPTION

This section provides a description of the integrated opti¬misation problem of airport ground operations combining the runway scheduling and ground movement problem similar to one introduced in [17].

A. Runway scheduling problem

The aim of runway scheduling in this paper is to find the optimal landing/take-off time of arriving/departing aircraft at the given runway minimising the delay and fuel burnt during waiting while respecting given safety constraints. In this paper, only take-off times are optimised whereas landing times are considered to be fixed, as from the practical point of view, it is easier to control taking-off aircraft still on the ground rather than airborne arriving aircraft.

The main constraint that limits the throughput of the runway is the minimum time interval between landing/taking-off aircraft. These minimum separations are due to wake vortices created by moving aircraft and in-flight separation constraints. In this paper, only separation due wake vortices is taken into account. The strength of wake vortices and thus separation depends on the aircraft type and is approximately proportional to its weight.

Let M = (A U D) be the set of total |M| = m arriving aircraft A and departing aircraft D. Let V (vi, vj) be the function to calculate the wake vortex separations from weight categories vi and vj of leading aircraft i and trailing aircraft j. The wake vortex separations used in this paper satisfy the triangle inequality V (vi, vj) + V (vj, ve) > V (vi, ve) for aircraft taking off in the order i, j, e and are given in Table I.

Let ri be the actual landing time for aircraft i E A and take¬off time for aircraft i E D. For arriving aircraft, ri is given. For departing aircraft, let di denote the time the departing

TABLE I: Separations in seconds between departing (D) and arriving (A) flights for weight classes: Heavy (H), Large (L), Small (S).

Trailing

A-H A-L A-S D-H D-L D-S

A-H 96 157 207 60 60 60

Leading A-L A-S 60

60 69

69 123

82 60

60 60

D-H 60 60 60 96 120 120

D-L 60 60 60 60 60 60

D-S 60 60 60 60 60 60

aircraft i E D arrived at the runway holding point. Aircraft i E D can take-off immediately, i.e. di = ri if there is enough time elapsed from landing/take-off time ri1 of the previous aircraft i  1 to comply with separation given by V (vi, vi1), otherwise, the departing aircraft i has to wait at the runway holding point until it is safe to take-off:

ri =

Then, the waiting time wi of the departing aircraft i E D is equal to wi = di  ri.

The objective of the runway scheduling is to minimise the total runway delay trwy and the total runway fuel fr wy burned by aircraft while waiting to take-off which depends on the delay wi and idle fuel flow 0vi specified for the weight category vi:

wi, (1)

wi • 0vi. (2)

The idle fuel flow 0vi corresponds to fuel flow from the Inter¬national Civil Aviation Organization (ICAO) engine database for 5 % of full power thrust of the representative aircraft, as explained in Section II-B.

B. Ground movement problem

The aim of the ground movement problem is to find routes and schedules for aircraft taxiing from runway to gate/stand and vice versa in a time and fuel efficient manner, respecting routes of other aircraft while preventing conflicts between them.

In this paper, due to the simple layout of the airport, the problem of finding routes is reduced to the shortest path problem. The shortest paths between each gate/stand and runway exit points are pre-calculated beforehand. Then, for each path optimised speed profiles are found by a specialized heuristic described in [30]. The speed profiles are optimised in a multi-objective manner, minimising taxi time and fuel consumption for the given route. The resulting non-dominated speed profiles are stored in a look-up table, with a separate table for each weight category vi and then retrieved during the on-line optimisation in order to save computational time.

Given the route of the aircraft and time needed to travel from origin to destination depends on the chosen speed profile, some delay may be added in order to prevent conflicts between taxiing aircraft. The conflicts between aircraft are solved by adding a small buffer time such that aircraft always maintain a safe time distance δ = 12 s between them (which corresponds to approximately 62 m at taxiing speed 10 knots).

Let yi be an integer representing the speed profile of aircraft i from the Pareto front of efficient speed profiles retrieved from the look-up table for the shortest route qi from the runway to the stand si for arriving aircraft i E A, or vice versa for departing aircraft i E D. We define a function T(qi, yi) which returns travel time of aircraft i taxiing on route qi for the given speed profile yi, including a delay to prevent taxiing conflicts.

In order to retrieve the fuel consumption of aircraft i of weight category vi, for route qi, using speed profile yi, a function F(qi, yi, vi) is defined. The function F(qi, yi, vi) is based on the method introduced in [30], which employs physics-based equations taking into account the acceleration force and rolling resistance to calculate thrust. The calculated thrust and aircraft engines are then mapped into corresponding fuel flows according to the ICAO emissions database.

Then, the objective of the ground movement problem is to minimise the total taxi time ttaxi and the total fuel ftaxi burned during ground movement:

ttaxi = T(yi), (3)

i=1

ftaxi = F(yi). (4)

i=1

C. Integrated optimisation problem

The runway scheduling and ground movement problem are combined into the integrated bi-objective optimisation problem with the following objective functions:

min g1 = ttaxi + trwy, (5)

min g2 = ftaxi + frwy, (6)

where g1 corresponds to the total time and g2 is the fuel consumption to be minimised. The decision variables for this optimisation problem are the pushback time for departing aircraft xi E 1-300, ..., 300} and the speed profile yi E 11, ... ,12} for all aircraft. The pushback time xi represents an integer number of seconds before/after the baseline time given as input flight schedule when the aircraft starts taxiing. The value of speed profile yi determines which solution from the pre-computed look-up table of non-dominated speed profiles is going to be selected, ranging from 1 representing the most time-efficient (fastest) speed profile to 12 standing for the most fuel efficient one. The number of speed profiles depends on the heuristic used for their generation, for details see [30].

Given the decision variables, the objective function values g1, g2 are determined as follows. Firstly, aircraft are considered sequentially according to their initial sequence specified by the input flight schedule. For each arriving aircraft i E A a

shortest route qi is retrieved from the look-up table between the runway and designated gate/stand si or in the opposite direction for departing aircraft i E D. The route qi, weight category vi and speed profile value yi determine which pre-computed speed profile is retrieved from the look-up table and used to schedule aircraft i along the route qi after all taxiing conflicts have been resolved. Then, the total taxi time ttaxi and the total fuel ftaxi are computed as stated in Section II-B. Based on taxi time T(yi), the runway holding point arrival time di is determined for each departing aircraft i E D. The delay wi is calculated as defined in section II-A and summed for all departing aircraft i E D to get the total runway delay trwy and subsequently the total fuel frwy. Finally, objectives g1, g2 are calculated according to 5 and 6.

D. Preferences based on costs

The bi-objective optimisation problem formulated in Sec¬tion II-C minimises two objective functions g1, g2 simulta¬neously. In practice, the total time g1 and fuel consumption g2 are associated with economic costs incurred by operations on the airport surface. For example, the fuel consumption g2 is connected to fuel cost defined by fuel price. Similarly, total time g1 incurs costs due to the fact, that each second of delay (i.e. unproductive time spent taxiing or waiting at runway) is an extra cost in terms of crew, maintenance, deprecation of aircraft, etc. Therefore, once costs related to g1, g2 are known, they can be used as weights during the optimisation, directing the search to a single cost-optimal solution. However, determining these cost, specifically cost of delay, is often difficult and only approximate as for example, crew wages or maintenance fees vary from airline to airline. Therefore, it might be more useful to guide the search to a region on the Pareto front containing solutions optimal for a certain cost range. We define a price vector c = [c1, ... , cn] which, in general, specifies cost for each one of n objectives. Furthermore, to include the uncertainty in costs, l boundary price vectors cB = [cB 1, ... , cB

n ] are defined. The number of boundary vectors depends on the DM. For example, DM can specify only upper bound for cost, both upper and lower bounds or different combinations in case of many objectives.

Then, each solution zj found during optimisation can be evaluated by function Ctotal(zj, c) in terms of total monetary cost, using price vector c:

Ctotal(zj, c) = En ck ' gk(zj). (7)

k =1

III. EMO FRAMEWORK WITH PREFERENCES

In order to optimise the objective functions g1, g2 of the integrated optimisation problem stated in Section II taking into account preferences given by costs, the following EMO framework is proposed in this section. Firstly, the DM is asked to provide a price vector c and l boundary price vectors cB as defined in Section II-D. Then, the search is divided into two parts:

1) A single objective optimisation search with aggre-gated objective function is performed by GA to quickly approach to the vicinity of the global Pareto front.

2) The best solutions from the single objective optimi

sation part are seeded into a preference-based EMO algorithm to reach RoI of the global Pareto front.

The single-objective optimisation is carried out by the GA minimising total costs Ctotal as defined in 7. The single-objective search is further accelerated by taking advantage of known preferences during the search process. For each aircraft i, speed profile is fixed to a value yi  {1,12} with minimum Ctotal. As a result, the search space is restricted to decision variable xi promising faster convergence, however sacrificing some of the optimality.

A. Preference-based NSGA-II

For the second phase, the EMO algorithm proposed in this paper, denoted as Preference-based NSGA-II (P-NSGA-II), is based on Light Beam Search EMO algorithm [23] which is a derivative of NSGA-II [27]. The initial population of random solutions is seeded with a specified number of best solutions (set to 10 in this paper) found by GA during the first phase. In the second phase, P-NSGA-II considers all decision variables xi and yi thus finding potentially better solutions. The P-NSGA-II performs the usual EMO operations: selection, crossover, mutation, replacement for the specified number of generations similar to original NSGA-II. In order to incorporate preferences into EMO algorithm, the replacement procedure which selects the solutions surviving to the next generation is modified. The original replacement procedure in NSGA-II favours non-dominated solutions with large crowding distance. In P-NSGA-II, the replacement is carried out with modified crowding distance to prefer solutions closer to RoI.

During replacement procedure, non-domination ranking is carried out for all solutions in the population. Then, the middle point and characteristic neighbours are determined as depicted on Fig. 1. The middle point ZC is a solution Zj with minimum cost defined by price vector C:

ZC = arg min (Ctotal(Zj, C)). (8)

The characteristic neighbour ZB, is a solution Zj for which the cost using the boundary price vector CB, is minimum:

ZB, = arg min (Ctotal(Zj, CB,)). (9)

Subsequently, for each solution Zj, belonging to the same non-dominated front, crowding distance cdj is calculated:

 if Zj is middle point ZC,

M + dj else if Zj outranks ZC,

1/Ctotal(Zj, C) otherwise.

If the solution Zj is the middle point, it is assigned an infinite crowding distance, so it is always included in the next generation. Other solutions are categorized according to their outranking relation to ZC. A solution Zj is said to outrank middle point ZC (denoted as ZjSZC) if mv(ZC, ZX) = 0, where mv is defined as follows:

mv(ZC, ZX) = card{k : gk(ZX)gk(ZC)  vk, k = 1,.. ., n} (10) Solutions outranking the middle point ZC are considered to be as good as ZC and form the outranking neighbourhood. As

Fig. 1: Preferred region on Pareto front.

all solutions in outranking neighbourhood belong to the same non-dominated front, if solution Zj is better than ZC in some objectives, then it must be worse in at least one other objective. Solution ZjSZC if this deterioration in objective k is not larger than veto threshold vk. The veto threshold vk for objective k is determined by characteristic neighbours:

vk = max{gk(ZB,), l = 1, ... , n}. (11)

Solutions Zj in the outranking neighbourhood are assigned M + dj crowding distance, where M is a big positive integer, set to M = 106 in this work and dj is the original crowding distance as defined by Deb [27]:

. (12)

gmax

k  gmin

Where gmax

k is the maximum and gmin

k the minimum value of the objective function gk found so far. The big integer M ensures that crowding distance cdj is large enough to guarantee inclusion into the next generation whereas dj causes the preference of solutions in less crowded areas, resulting in a uniformly distributed solutions across the characteristic neighbourhood.

For remaining solutions, the crowding distance cdj is inversely related to Ctotal(Zj, C). As a result, solutions Zj with smaller Ctotal(Zj, C) and thus closer to the RoI are assigned larger crowding distance giving them larger chance to be included in the next generation.

IV. COMPUTATIONAL RESULTS AND DISCUSSION

A. Experimental setup

The proposed EMO framework was tested on a dataset of real arrival and departure flights on Doha International Airport (DOH) which was the largest airport in Qatar and a hub airport for Qatar Airways until the new Hamad International airport was completed in late April 2014. DOH airport has one runway and 55 stands. The data were recorded on 16th March 2014 and divided into two instances doh1 representing medium traffic conditions and doh2 for high traffic conditions. The instance doh1 includes 96 flights between 17:00 and 21:00 UTC from which 50 are arrivals and 46 departures. The instance doh2

TABLE II: Specifications of the representative aircraft.

Learjet 35A Airbus A320 Airbus A333

Take-off weight 8300 kg 78000 kg 230000 kg

Engines TFE731-2-2B CMF56-5-A1 CF6-80E1A2

Number of engines 2 2 2

Rated output R 2x15.6 kN 2x111.2 kN 2x287 kN

Rolling resistance 1221 N 11.48 kN 33.84 kN

Fuel flow at 7% R 0.024 kg-s1 0.101 kg-s1 0.228 kg-s1

Fuel flow at 30% R 0.067 kg-s1 0.291 kg-s1 0.724 kg-s1

TABLE III: Average IE indicator for 30 runs.

NSGA-II P-NSGA-II P-NSGA-II with seed

doh1 1.061 1.053 1.016

doh2 1.219 1.207 1.040

Fig. 2: Pareto fronts for doh1.

consists of 84 flights between 21:00 and 23:00 UTC from which there are 27 arrivals and 57 departures. The data provided specified landing/pushback times and gates/runway exits for each flight.

The aircraft have been divided into 3 groups according their wake vortex separation requirements as defined in [31]. For each category, a representative aircraft is designated and its specifications are used during the fuel consumption calculation. The specifications are summarized in Table II.

The computational experiments were performed on a com¬puter with an Intel i3-2120 processor and 3.16 GB of RAM, running Linux. The EMO framework is implemented using the Inspyred package for Python [32]. The termination criteria for EMO framework was set to 900 seconds, of which 600 seconds is allocated to GA and 300 seconds to P-NSGA-II. Based on initial experiments, the number of individuals in population for GA was set to 200 and 50 for P-NSGA-II. The price vector c is equal to [0.469, 0.71], where the first element corresponds to costs of delay as defined in [17] and the second element is the jet fuel price (as of 14.1.2014), both in Euro. The boundary vector cB1 = [1.2  0.469, 0.8  0.71] and cB2[0.8  0.469,1.2  0.71].

B. Computational results

The performance of the proposed two-phase EMO frame¬work (P-NSGA-II with seed) was compared to original NSGA-II and P-NSGA-II without the seed from single-objective GA. Fig. 2 shows the best Pareto fronts from 30 runs of the algorithms for doh1 instance. As can be seen, single-objective GA was able to find better solutions than NSGA-II and comparable with P-NSGA-II alone. Then, P-NSGA-II could further improve the solutions from the initial seed, focusing the search on RoI.

The multiplicative unary epsilon indicator IE is used as the performance index, expressing the distance of the resulting Pareto front to the global Pareto front, which was constructed by considering all solutions found during the experiments and leaving only non-dominated ones. Average IE indicator values for 30 runs of the algorithms are given in Table III.



Fig. 3: Averaged convergence for 30 runs for doh1 instance.

The results of IE show that P-NSGA-II alone could only marginally improve the performance compared to NSGA-II. However, adding seed to P-NSGA-II significantly helped the algorithm to find better solutions. The improving effect is even more evident for doh2 instance with high traffic conditions. We can hypothesize that beneficial impact of initial single-objective GA on the overall performance of the EMO framework will increase in case of many (i.e. more than 3) objectives as indicated in [22], due to the fact that many objectives resulting in high number of non-dominated solutions deteriorate the performance of EMO algorithms.

The convergence during the evolution for the doh1 is documented in Fig. 3. Single-objective GA in the first phase of the search quickly finds good solutions, helping the P-NSGA-II in the second phase obtain good values of I, Furthermore, fixing speed profiles yi for GA results in better values of IE for the initial population, effectively accelerating the search.

V. CONCLUSION

In this paper, a multi-objective evolutionary framework tak¬ing into account preferences in an a-priori manner is proposed to solve integrated optimisation problem of runway scheduling and ground movement on airport surface. The preference was incorporated in the first phase to transform the original multi-objective problem to a single-objective one to quickly approach

the region of interest. Then, in the second phase a modified crowding distance in NSGA-II was applied to finally arrive to Pareto front of cost-efficient solutions. The preliminary computational experiments conducted on real-world data from a major Asian airport showed promising results and indicate that the proposed approach could be suitable as an optimisation framework for decision support at the airport. The evolutionary framework utilising the preferences given as costs associated with objectives successfully accelerated the search. Moreover, the input values of costs enabled to control the extent of the resulting region and include some uncertainty about the prices. The incorporation of preferences helps the decision maker to focus his attention on only cost-effective solutions in a shorter computational time. For the future research, the approach needs to be validated on a more diverse set of problem instances. Also, the idea of including uncertainty into preference information deserves more attention. As an example, the uncertainty could be expressed in terms of fuzzy values.

REFERENCES

[1] ICAO. (2014) Annual Report of the ICAO Council: 2013 The World of Air Transport. [Online]. Available: http://www.icao.int/ annual-report-2013/Pages/the-world-of-air-transport-in-2013.aspx

[2] L. Bianco, P. Dell’Olmo, and S. Giordani, “Scheduling models for air traffic control in terminal areas,” Journal of Scheduling, vol. 9, no. 3, pp. 223–253, 2006. [Online]. Available: http: //dx.doi.org/10.1007/s10951-006-6779-7

[3] M. Sam`a, A. D’Ariano, and D. Pacciarelli, “Rolling horizon approach for aircraft scheduling in the terminal control area of busy airports,” Transportation Research Part E: Logistics and Transportation Review, vol. 60, no. 0, pp. 140–155, 2013. [Online]. Available: http://www.sciencedirect.com/science/article/pii/S136655451300118X

[4] J. A. Atkin, E. K. Burke, and S. Ravizza, “The airport ground movement

problem: Past and current research and future directions,” Proceedings of the 4th International Conference on Research in Air Transportation (ICRAT), Budapest, Hungary, pp. 131–138, 2010.

[5] H. Balakrishnan and B. Chandran, “Scheduling aircraft landings under constrained position shifting,” in AIAA Guidance, Navigation, and Control Conference and Exhibit, Keystone, CO, 2006.

[6] J. A. D. Atkin, E. K. Burke, J. S. Greenwood, and D. Reeson, “Hybrid Metaheuristics to Aid Runway Scheduling at London Heathrow Airport,” Transportation Science, vol. 41, no. 1, pp. 90–106, 2007.

[7] X.-B. Hu and E. Di Paolo, “Binary-Representation-Based Genetic Algorithm for Aircraft Arrival Sequencing and Scheduling,” Intelligent Transportation Systems, IEEE Transactions on, vol. 9, no. 2, pp. 301– 310, June 2008.

[8] J. Bennell, M. Mesgarpour, and C. Potts, “Airport runway scheduling,” Annals of Operations Research, vol. 204, no. 1, pp. 249–270, 2013. [Online]. Available: http://dx.doi.org/10.1007/s10479-012-1268-1

[9] A. Mar´ın and E. Codina, “Network design: taxi planning,” Annals of Operations Research, vol. 157, no. 1, pp. 135–151, 2008.

[10] P. C. Roling and H. G. Visser, “Optimal Airport Surface Traffic Planning Using Mixed-Integer Linear Programming,” International Journal of Aerospace Engineering, vol. vol. 2008, p. 11, 2008.

[11] S. Ravizza, J. A. Atkin, and E. K. Burke, “A more realistic approach for airport ground movement optimisation with stand holding,” Journal of Scheduling, pp. 1–14, 2013.

[12] C. Lesire, “An Iterative A* Algorithm for Planning of Airport Ground Movements,” 19th European Conference on Artificial Intelligence (ECAI)/6th Conference on Prestigious Applications of Intelligent Sys¬tems (PAIS), Lisbon, Portugal, August 16-20, 2010.

[13] S. Ravizza, J. Chen, J. A. Atkin, E. K. Burke, and P. Stewart, “The trade-off between taxi time and fuel consumption in airport ground movement,” Public Transport, vol. 5, no. 1-2, pp. 25–40, 2013.

[14] R. Deau, J. Gotteland, and N. Durand, “Airport surface management and runways scheduling,” in Proceedings of the 8th USA/Europe Air Traffic Management Research and Development Seminar, Napa, CA, USA, 2009.

[15] G. Clare and A. Richards, “Optimization of Taxiway Routing and Runway Scheduling,” Intelligent Transportation Systems, IEEE Trans¬actions on, vol. 12, no. 4, pp. 1000–1013, Dec 2011.

[16] M. Frankovich and D. Bertsimas, “Air Traffic Flow Management at Airports: A Unified Optimization Approach,” in Tenth USA/EUROPE Air Traffic Management Research & Development Seminar, June 2013.

[17] M. Weiszer, J. Chen, and G. Locatelli, “An integrated optimisation approach to airport ground operations to foster sustainability in the aviation sector,” Applied Energy, 2015, in press. [Online]. Available: http://dx.doi.org/10.1016/j.apenergy.2015.04.039.

[18] K. Deb, J. Sundar, N. Udaya Bhaskara Rao, and S. Chaudhuri, “Refer¬ence point based multi-objective optimization using evolutionary algo¬rithms,” International Journal of Computational Intelligence Research, vol. 2, no. 3, pp. 273–286, 2006.

[19] L. Ben Said, S. Bechikh, and K. Gh´edira, “The r-dominance: a new dominance relation for interactive evolutionary multicriteria decision making,” Evolutionary Computation, IEEE Transactions on, vol. 14, no. 5, pp. 801–818, 2010.

[20] J. Molina, L. V. Santana, A. G. Hern´andez-D´ıaz, C. A. Coello Coello, and R. Caballero, “g-dominance: Reference point based dominance for multiobjective metaheuristics,” European Journal of Operational Research, vol. 197, no. 2, pp. 685–692, 2009.

[21] L. Thiele, K. Miettinen, P. J. Korhonen, and J. Molina, “A preference-based evolutionary algorithm for multi-objective optimization,” Evolu¬tionary Computation, vol. 17, no. 3, pp. 411–436, 2009.

[22] A. L´opez-Jaimes and C. A. C. Coello, “Including preferences into a multiobjective evolutionary algorithm to deal with many-objective engineering optimization problems,” Information Sciences, vol. 277, no. 0, pp. 1–20, 2014. [Online]. Available: http: //www.sciencedirect.com/science/article/pii/S0020025514004708

[23] K. Deb and A. Kumar, “Light beam search based multi-objective opti¬mization using evolutionary algorithms,” in Evolutionary Computation, 2007. CEC 2007. IEEE Congress on. IEEE, 2007, pp. 2125–2132.

[24] S. Huber, M. J. Geiger, M. Sevaux et al., “Interactive approach to the Inventory Routing Problem: computational speedup through focused search,” in Proceedings of the Logistikmanagement 2013 conference, 2013.

[25] T. Friedrich, T. Kroeger, and F. Neumann, “Weighted preferences in evolutionary multi-objective optimization,” in AI 2011: Advances in Artificial Intelligence. Springer, 2011, pp. 291–300.

[26] D. Cvetkovic and I. Parmee, “Preferences and their application in evolutionary multiobjective optimization,” Evolutionary Computation, IEEE Transactions on, vol. 6, no. 1, pp. 42–57, Feb 2002.

[27] K. Deb, A. Pratap, S. Agarwal, and T. Meyarivan, “A fast and elitist multiobjective genetic algorithm: NSGA-II,” Evolutionary Computa¬tion, IEEE Transactions on, vol. 6, no. 2, pp. 182–197, Apr. 2002.

[28] J. Branke and K. Deb, “Integrating user preferences into evolutionary multi-objective optimization,” in Knowledge incorporation in evolution¬ary computation. Springer, 2005, pp. 461–477.

[29] A. Jaszkiewicz and R. Słowi´nski, “The ‘Light Beam Search’approach– an overview of methodology applications,” European Journal of Oper¬ational Research, vol. 113, no. 2, pp. 300–314, 1999.

[30] M. Weiszer, J. Chen, S. Ravizza, J. Atkin, and P. Stewart, “A heuristic approach to greener airport ground movement,” July 2014, pp. 3280– 3286.

[31] ICAO, Air Traffic Management: Procedures for Air Navigation Services, 2007, 15th edition. Doc 4444 ATM/501.

[32] A. Garrett, “Inspyred,” 2012, (Version 1.0) [software]. Inspired Intelli¬gence Initiative. Retrieved from http://inspyred.github.com.

Proceedings of the World Congress on Engineering 2009 Vol I

WCE 2009, July 1 - 3, 2009, London, U.K.

Software Fault Proneness Prediction Using

Support Vector Machines

Yogesh Singh, Arvinder Kaur, Ruchika Malhotra

Abstract— Empirical validation of software metrics to predict quality using machine learning methods is important to ensure their practical relevance in the software organizations. In this paper, we build a Support Vector Machine (SVM) model to find the relationship between object-oriented metrics given by Chidamber and Kemerer and fault proneness. The proposed model is empirically evaluated using public domain KC1 NASA data set. The performance of the SVM method was evaluated by Receiver Operating Characteristic (ROC) analysis. Based on these results, it is reasonable to claim that such models could help for planning and performing testing by focusing resources on fault-prone parts of the design and code. Thus, the study shows that SVM method may also be used in constructing software quality models.

Keywords: Metrics, Object-oriented, Software Quality, Empirical validation, Fault prediction, Support vector machine, Receiver Operating Characteristics analysis

I. INTRODUCTION

As the complexity and the constraints under which the software is developed are increasing, it is difficult to produce software without faults. One way to deal with this problem is to predict important software quality attributes such as fault-proneness, effort, testability, maintainability, and reliability during early phases of software development. The software metrics [4, 9, 10, 14, 15, 24, 25, 29-32, 41] may be used in predicting these quality attributes.

Manuscript received March 21, 2009.

Prof. Yogesh Singh is with Guru Gobind Singh Indraprastha

University, Delhi, India (email: ys66@rediffmail.com)

Dr. Arvinder Kaur is with Guru Gobind Singh Indraprastha

University, Delhi, India (e-mail:

arvinderkaurtakkar@yahoo.com.)

Ruchika Malhotra (Corresponding Author phone: 91-011-26431421) is with Guru Gobind Singh Indraprastha University, Delhi, India (email: ruchikamalhotra2004@yahoo.com)

The behaviour of several quantitative models ranging from using simple linear discriminant analysis to more complex logistic regression, decision tree, and SVM have been proposed. The regression and machine learning approaches are inherently different, raising the question to analyze the performance of these methods.

Several empirical studies have been carried out to predict the fault proneness models such as [1, 2, 5, 7, 11, 12, 14, 16, 20, 21, 23, 27, 34, 35, 40, 44]. There is a need to empirically validate the machine learning methods in predicting software quality attributes. Therefore, more data-based empirical studies that can be used to verify the capability of machine learning methods are needed. The evidence gathered through these empirical studies is considered to be the strongest support for testing a given hypothesis.

Thus, there is a need for both 1) empirically validating the results of machine leaning methods such as SVM and 2) finding the relation between OO metrics given by Chidamber and Kemerer [15] and fault proneness models. Now we briefly describe the work done in this study. In this paper, we investigate the following issue:

• Are the fault proneness models predicted using SVM method feasible and adaptable?

• How accurately and precisely do the OO metrics predict faults?

In order to perform the analysis we validate the performance of the SVM method using public domain KC1 NASA data set [42]. The 145 classes in this data were developed using C++ language. The contributions of this paper are summarized as follows: First, we performed the analysis of public domain NASA data set [42], therefore analyzing valuable data in an important area where empirical studies and data are limited. Second, we applied SVM method to predict the effect of OO metrics on fault proneness. To the best of our knowledge,

there has been no such

ISBN: 978-988-17012-5-1 WCE 2009

Proceedings of the World Congress on Engineering 2009 Vol I

WCE 2009, July 1 - 3, 2009, London, U.K.

previous research using SVM method to predict software fault proneness. The proposed results showed that SVM method predict faulty classes with high accuracy. However, since our analysis is based on only one data set, this study should be replicated on different data sets to generalize our findings.

The paper is organized as follows: Section 2 summarizes the metrics studied, and describes sources from which data is collected. Section 3 presents the overview of the SVM method. The results of the study are given in section 4. The model is evaluated in section 5. Conclusions of the research are presented in section 6.

II. RESEARCH BACKGROUND

In this section, we present the summary of metrics studied in this paper (Section A), and empirical data collection (Section B).

A. Dependent and Independent Variables

The binary dependent variable in our study is fault proneness. The goal of our study is to explore empirically the relationship between OO metrics

and fault proneness. Fault proneness is defined as the probability of fault detection in a class [2]. We use machine learning methods to predict the probability of fault proneness. Our dependent variable will be predicted based on the faults found during software development life cycle. The metrics given by [15] are summarized in Table 1.

B. Empirical Data Collection

This study makes use of the public domain data set KC1 from the NASA Metrics Data Program [42]. The data in KC1 was collected from a storage management system for receiving/processing ground data, which was implemented in the C++ programming language. This system consists of 145 classes that comprise of 2107 methods, with 40K lines of code. KC1 provides both class-level and method-level static metrics. At the method level, 21 software product metrics based on product’s complexity, size and vocabulary are given. Five types of defects such as the number of defects and density of defects are also given. At the class level, values of 10 metrics are computed including seven metrics given by Chidamber and Kemerer [15]. These seven OO metrics are taken in our study (see Table 1) for analyses.

TABLE 1

METRICS STUDIED

Metric Definition

Coupling between Objects (CBO) CBO for a class is count of the number of other classes to which it is coupled and vice versa.

Lack of Cohesion (LCOM) For each data field in a class, the percentage of the methods in the class using that data field; the percentages are averaged then subtracted from 100%.

Number of Children (NOC) The NOC is the number of immediate subclasses of a class in a hierarchy.

Depth of Inheritance (DIT) The depth of a class within the inheritance hierarchy is the maximum number of steps from the class node to the root of the tree and is measured by the number of ancestor classes.

Weighted Methods per Class (WMC) A count of methods implemented within a class.

Response for a Class (RFC) A count of methods implemented within a class plus the number of methods accessible to an object class due to inheritance.

Source Lines Of Code

(SLOC) It counts the lines of code.

ISBN: 978-988-17012-5-1 WCE 2009

Proceedings of the World Congress on Engineering 2009 Vol I

WCE 2009, July 1 - 3, 2009, London, U.K.

III. MODEL PREDICTION USING SUPPORT VECTOR MACHINE (SVM) METHOD

SVM are useful tools for performing data classification, and have been successfully used in applications such as face identification, medical diagnosis, text classification [43], pattern recognition [13], chinese character classification [45], and identification of organisms [33]. SVM constructs an N-dimensional hyperplane that optimally separates the data set into two categories. The purpose of SVM modeling is to find the optimal hyperplane that separates clusters of vector in such a way that cases with one category of the dependent variable on one side of the plane and the cases with the other category on the other side of the plane [37]. The support vectors are the vectors near the hyperplane. The SVM modeling finds the hyperpalne that is oriented so that the margin between the support vectors is maximized. When the points are separated by a nonlinear region, SVM handles this by using a kernel function inorder to map the data into a different space when a hyperplane can be used to do the separation. Details on SVM can be found in [17, 18].

The recommended kernel function is the Radial basis Function (RBF) [37]. Thus, we used RBF function in SVM modeling to predict faulty classes in this study. The RBF kernel maps non-linearly data into a higher dimensional space, so it can handle non linear relationships between the dependent and the independent variables. Figure 1 shows the RBF kernel. One category of the

Figure 1: Radial basis function

other as triangles. The shaded circles and rectangles are support vectors.

dependent variable is shown as rectangles and the

Given a set of (xi, yi), ,(xm, ym) and

yi {1,+1} training samples. αi =(i=1, ,m) is

a lagrangian multipliers. K(xi,yi) is called a kernel function and b is a bias. The discriminant function D of two class SVM is given below [45]:

D(x) ==yiαiK(xi, x) + b (1)

i 1

Then an input pattern x is classified as [45]:

+1 if D(x)>0 x =  1 if D(x) < 0  (2)

The performance of the models predicted was evaluated using sensitivity, specificity, precision, completeness, and Area Under the Curve (AUC). Details on these measures can be found in [19]. The ROC curve, which is defined as a plot of sensitivity on the y-coordinate versus its 1-specificity on the x coordinate, is an effective method of evaluating the quality or performance of the predicted models [22].

IV. ANALYSIS RESULTS

This section presents the analysis results, following the procedure described in Section 3. The results of the model predicted and evaluated are presented.

Results of prediction model: Table 2 shows the sensitivity, specificity, precision, completeness, and cutoff point of each metric and the model predicted. SLOC and CBO metrics have the highest values of the sensitivity and completeness. However, in case of NOC metric all the classes were predicted to be non faulty hence the results are not shown, as testing all the classes will be a high waste of the testing resources.

The model was applied to 145 classes and Table 3 presents the results of correctness of the fault proneness model predicted. As shown in Table 2, the cut off point for the model build to predict fault proneness is 0.44. Out of 59 classes, actually fault prone, 45 classes were predicted to be fault prone. The sensitivity of the model is 76.27 percent. Similarly, 70 out of 86 classes were predicted not to be fault prone. Thus, the specificity of the model is 81.39 percent. The completeness of the model is

85.66 percent. Thus,

ISBN: 978-988-17012-5-1 WCE 2009

Proceedings of the World Congress on Engineering 2009 Vol I

WCE 2009, July 1 - 3, 2009, London, U.K.

the accuracy of the model is very high.

V. MODEL EVALUATION USING ROC ANALYSIS

The accuracy of the models predicted is somewhat optimistic since the models are applied on same data set from which they are derived. To predict accuracy of the model it should be applied on different data sets thus we performed 10-cross validation of the SVM model. Details on cross validation can be found in [39]. Table 4 summarizes the results of 10-cross validation of models using the SVM method. The AUC of the model predicted is 0.89.

In line with other predictive models, likewise findings of this study need to be externally validated. Although regression analysis is widely

used method in literature, our results show that performance of the SVM model is good. In a previous study, we validated SVM using open source data set [38]. In [38], results also showed good performance of the SVM method. Therefore, it appears that the model predicted using machine-learning methods might lead to development of optimum software quality models for predicting fault prone classes.

Planning and resource allocating for inspection and testing is difficult and it is usually done on empirical basis. The model predicted in the above section could be of great help for planning and executing testing activities. To illustrate how the prediction model can be applied in practice, consider Figure 2. The values for the predicted fault proneness were taken from the results of validation of the models. On X-axis, we plot

TABLE 2

SENSITIVITY, SPECIFICITY, PRECISION, AND COMPLETENESS

Metric Sensitivity Specificity Precision Completeness Cutoff

CBO 76.27 68.6 71.7 82.24 0.40

WMC 55.93 65.1 61.37 71.18 0.41

RFC 52.54 66.27 60.68 67.13 0.37

SLOC 62.7 75.58 70.34 76.32 0.32

LCOM 59.32 63.95 62.75 65.88 0.33

NOC - - - - -

DIT 71.18 29 46.2 78.34 0.38

Model IV 76.27 81.39 78.62 85.66 0.44

TABLE 3

PREDICTED CORRECTNESS OF MODEL

Predicted

Po<=0.43 Po>0.43

Observed Not faulty Faulty

Not faulty 70 16

Faulty 14(92) 45(550)

TABLE 4

. RESULTS OF 10-CROSS VALIDATION OF MODELS

SUPPORT VECTOR MACHINE

Model

Cutoff point 0.45

Sensitivity 69.49

Specificity 82.55

Precision 77.24

Completeness 79.59

AUC 0.89

ISBN: 978-988-17012-5-1 WCE 2009

Proceedings of the World Congress on Engineering 2009 Vol I WCE 2009, July 1 - 3, 2009, London, U.K.

the probability of classes sorted in increasing order of their cutoff points. On Y-axis each column presents two lines. The first and second line shows the sensitivity and completeness of the predicted model, respectively. For example, in Figure 2 at X=0, the sensitivity of the model is 100% and the completeness is 100%.

VI. CONCLUSIONS

The goal of our research is to empirically analyze the performance of the SVM method. We find the individual and combined effect of each metric on fault proneness of classes. Based on public domain NASA data set KC1, we empirically analyzed the performance of the SVM method using ROC analysis.

We analyzed OO design metrics given by Chidamber and Kemerer. Our main results are as follows:

• The CBO, RFC, and SLOC metrics

were found to be related to fault proneness. NOC and DIT metric was not found to be significantly related to fault proneness. Thus, the results of machine learning method (SVM) show that usefulness of NOC and DIT metric is poor.

• The SVM method yielded good AUC

using ROC analysis. This study confirms that construction of the SVM models is feasible, adaptable to OO systems, and useful in predicting fault prone classes. While research continues, practitioners and researchers may apply machine learning methods for constructing the model to predict faulty classes.

As in all empirical studies, the relationship we established is valid only for certain population of systems. In this case, the authors can roughly characterize this population as “object-oriented, large-sized systems.”

We plan to replicate our study to predict the models based on machine learning algorithms such as genetic algorithms. We will also focus on cost benefit analysis of models that will help to determine whether a given fault proneness model would be economically viable.

Figure 2: Sensitivity and Completeness of the

Model

REFERENCES

[1] Aggarwal K.K., Singh Y., Kaur A., Malhotra R. (2007). Application of Artificial Neural Network for Predicting Fault Proneness Models, International Conference on Information Systems, Technology and. Management (ICISTM 2007), March 12-13, New Delhi, India.

[2] Aggarwal K.K., Singh Y., Kaur A., Malhotra R. (2008). Empirical Analysis for Investigating the Effect of Object-Oriented Metrics on Fault Proneness: A Replicated Case Study, Forthcoming in Software Process Improvement and Practice, Wiley.

[3] Aggarwal K.K., Singh Y., Kaur A., Malhotra R. (2006). Empirical study of object-oriented metrics. Journal of Object Technology, 5(8), 149-173.

[4] Aggarwal K.K., Singh Y., Kaur A., Malhotra R. (2005). Software reuse metrics for object-oriented systems. In Proceedings of the Third ACIS Int’l Conference on Software Engineering Research, Management and Applications (SERA ’05), 48-55.

[5] Aggarwal K.K., Singh Y., Kaur A., Malhotra R. (2006). Investigating the Effect of Coupling Metrics on Fault Proneness in Object-Oriented Systems. Software Quality Professional, 8(4), 4-16.

[6] Barnett V., Price T. (1995). Outliers in Statistical Data. John Wiley & Sons.

[7] Basili,V., Briand, L., and Melo, W. (1996). A validation of object-oriented design metrics as quality indicators. IEEE Transactions on Software Engineering, 22(10), 751¬761.

[8] Binkley, A., and Schach, S. (1998). Validation of the coupling dependency metric as a risk predictor. In Proceedings of the International Conference on Software Engineering, 452-455.

[9] Briand, L., Daly, W., and Wust J. (1998). Unified framework for cohesion measurement in object-oriented systems. Empirical Software Engineering, 3(1), 65-117.

[10] Briand, L., Daly, W., and Wust J. (1999). A unified framework for coupling measurement in object-oriented systems. IEEE Transactions on Software Engineering , 25(1), 91-121.

[11] Briand, L., Daly, W., and Wust J. (2000). Exploring the relationships between design measures and software quality. Journal of Systems and Software, 51(3), 245-273.

[12] Briand L., Wüst J., Lounis H. (2001). Replicated Case Studies for Investigating Quality Factors in Object-Oriented Designs, Empirical Software Engineering: An International Journal , 6(1), 11-58.

[13] Burges, C. (1998). A tutorial on support vector machines for pattern recognition. Data Mining and Knowledge Discovery 2, 121–167.

ISBN: 978-988-17012-5-1 WCE 2009

Proceedings of the World Congress on Engineering 2009 Vol I WCE 2009, July 1 - 3, 2009, London, U.K.

[14] Cartwright, M., and Shepperd, M. (1999). An empirical investigation of an object-oriented software system. IEEE Transactions of Software Engineering, 26(8), 786-796.

[15] Chidamber, S., and Kamerer, C. (1994). A metrics suite for object-oriented design. IEEE Transactions on Software Engineering , 20(6), 476-493.

[16] Chidamber, S., Darcy, D., and Kemerer, C. (1998). Managerial use of metrics for object-oriented software: An exploratory analysis. IEEE Transactions on Software Engineering, 24(8), 629-639.

[17] Cortes, C., Vapnik, V. (1995). Support-vector networks. Machine Learning 20, 273–297.

[18] Cristianini, N., Shawe-Taylor, J. (2000). An Introduction to Support Vector Machines and Other Kernel-based Learning Methods. Cambridge University Press, Cambridge, UK.

[19] El Emam, K., Benlarbi, S., Goel, N. And Rai, S. 1999. A Validation of Object-Oriented Metrics, Technical Report ERB-1063, NRC.

[20] El Emam, K., Benlarbi, S., Goel, N., and Rai, S. (2001). The Confounding Effect of Class Size on The Validity of Object-Oriented Metrics. IEEE Transactions on Software Engineering, 27(7), 630-650.

[21] Gyimothy, T., Ferenc, R., Siket, I. (2005). Empirical validation of object-oriented metrics on open source software for fault prediction, IEEE Trans. Software Engineering, 31 (10), 897 – 910.

[22] Hanley, J., McNeil, BJ. (1982). The meaning and use of the area under a Receiver Operating Characteristic ROC curve. Radiology, 143: 29-36.

[23] Harrison, R., Counsell, S. J., and Nithi, R.V. (1998). An evaluation of MOOD set of object-oriented software metrics. IEEE Transactions on Software Engineering, 24(6), 491-496.

[24] Henderson-Sellers, B. (1996). Object-oriented metrics, measures of complexity. Englewood Cliffs, N.J.: Prentice Hall.

[25] Hitz, M., and Montazeri, B. (1995). Measuring coupling and cohesion in object-oriented systems. In Proceedings of the International Symposium on Applied Corporate Computing, Monterrey, Mexico.

[26] Hosmer, D., and Lemeshow, S. (1989). Applied logistic regression. New York: John Wiley and Sons.

[27] Khoshgaftaar, T.M., Allen, E.D., Hudepohl, J.P, Aud, S.J. (1997). Application of neural networks to software quality modeling of a very large telecommunications system," IEEE Transactions on Neural Networks, 8(4), 902-909.

[28] Kothari, C. R. (2004). Research Methodology. Methods and Techniques. New Delhi: New Age International Limited.

[29] Lake, A., and Cook, C. (1994). Use of factor analysis to develop OOP software complexity metrics. In Proceedings of the 6th Annual Oregon Workshop on Software Metrics, Silver Falls, Oregon.

[30] Lee, Y., Liang, B., Wu, S., and Wang, F. (1995). Measuring the coupling and cohesion of an object-oriented program based on information flow. In Proceedings of the International Conference on Software Quality, Maribor, Slovenia.

[31] Li, W., and Henry, S. (1993). Object-oriented metrics that predict maintainability. Journal of Systems and Software, 23(2), 111-122.

[32] Lorenz, M., and Kidd, J. (1994). Object-oriented software metrics. Englewood Cliffs, N.J.: Prentice-Hall.

[33] Morris, C., Autret, A., Boddy, L. (2001). Support vector machines for identifying organisms-a comparison with strongly partitioned radial basis function networks. Ecological Modeling 146, 57–67.

[34] Olague, H., Etzkorn, L., Gholston, S., and Quattlebaum, S. (2007). Empirical Validation of Three Software Metrics Suites to Predict Fault-Proneness of Object-Oriented Classes Developed Using Highly Iterative or Agile Software Development Processes. IEEE Transactions on software Engineering, 33(8), 402-419.

[35] Pai, G. (2007). Empirical analysis of Software Fault Content and Fault Proneness Using Bayesian Methods, IEEE Transactions on software Engineering, 33(10), 675-686.

[36] Porter, A., Selly, R. (1990). Empirically guided Software Development using Metric-Based Classification Trees, IEEE Software, 7(2), 46-54.

[37] Sherrod, P. (2003) DTreg Predictive Modeling Software.

[38] Singh, Y., Arvinder, K., Malhotra, R. (2009). Application of Support Vector Machine to Predict Fault Prone Classes, ACM SIGSOFT Software Engineering Notes, 34( 9), 1-6.

[39] Stone, M. (1974). Cross-validatory choice and assessment of statistical predictions. J. Royal Stat. Soc., 36, 111-147.

[40] Tang, M.H, Kao,, M.H., and Chen, M.H. (1999). An Empirical Study on Object-Oriented Metrics, In Proceedings of Metrics, 242-249.

[41] Tegarden, D., Sheetz, S., and Monarchi, D. (1995). A software complexity model of object-oriented systems. Decision Support Systems 13 (3-4), 241-262.

[42] www.mdp.ivv.nasa.gov, NASA Metrics data Repository.

[43] Wang, X., Bi, D., and Wang, S. (2007). Fault recognition with Labeled multi-category’, Third conference on Natural Computation, Haikou, China.

[44] Yuming, Z. and Hareton, L. (2006). Empirical analysis of Object-Oriented Design Metrics for predicting high severity faults. IEEE Transactions on Software Engineering, 32(10), 771-784.

[45] Zhao, L., Takagi, N. (2007). An application of Support vector machines to Chinese character classification problem. IEEE International Conference on systems, Man and Cybernetics, Montreal.

ISBN: 978-988-17012-5-1 WCE 2009

Communications of the

A I S

ssociation for nformation ystems

Panel Report DOI: 10.17705/1CAIS.043XX ISSN: 1529-3181

How do Machine Learning, Robotic Process

Automation, and Blockchains Affect the Human Factor

in Business Process Management?

Wirtschaftsuniversität Wien, Vienna

Austria

jan.mendling@wu.ac.at

Signavio IBM Research

Germany USA

Vrije Universiteit Amsterdam Data61, CSIRO

The Netherlands Australia

This paper summarizes a panel discussion at the 15th International Conference on Business Process Management. The panel discussed to what extent the emergence of recent technologies including machine learning, robotic process automation, and blockchain will reduce the human factor in business process management. The panel discussion took place on 14 September, 2017, at the Universitat Politècnica de Catalunya in Barcelona, Spain. Jan Mendling served as a chair; Gero Decker, Richard Hull, Hajo Reijers, and Ingo Weber participated as panelists. The discussions emphasized the impact of emerging technologies at the task level and the coordination level. The major challenges that the panel identified relate to employment, technology acceptance, ethics, customer experience, job design, social integration, and regulation.

Business Process Management, Process Automation, Artificial Intelligence, Machine Learning, Robotic Process Automation, Blockchain.

[Department statements, if appropriate, will be added by the editors. Teaching cases and panel reports will have a statement, which is also added by the editors.]

[Note: this page has no footnotes.]

This manuscript underwent peer review. It was received 12/09/2017 and was with the authors for 1 month for 1 revision. Christoph Peters served as Associate Editor.

Volume 43 Paper XXX pp. 1 – 23 June 2018

2 How do Machine Learning, Robotic Process Automation and Blockchains Affect the Human Factor in BPM?

1 Introduction

The business process management (BPM) discipline investigates methods and techniques to organize business processes in an efficient and effective manner (Dumas, La Rosa, Mendling, & Reijers, 2013). A key idea of BPM involves improving business processes by redesigning information systems to best support the people who are working in the process. Indeed, many early office automation systems (Hirschheim, 1985), workflow systems (van der Aalst & van Hee, 2004), and various more recent process-aware information systems (Dumas, van der Aalst, & ter Hofstede, 2005)—which researchers often subsume under the term BPM systems (dumas et al., 2013)—all focus on this idea. Such systems hold and provide information to workers, schedule and coordinate specific pieces of work, and support decisions on how to best proceed.

Recent advancements in the area of artificial intelligence, machine learning, cryptography, and distributed systems have provided the foundations for new technologies, including robotic process automation (Aguirre & Rodriguez, 2017), chatbots (Shawar & Atwell, 2007), self-driving cars (Daily, Medasani, Behringer, & Trivedi, 2017), smart objects (Beverungen, Müller, Matzner, Mendling, & vom Brocke, 2017), blockchains (Nakamoto, 2008), and the Internet of things (Atzori, Iera, Morabito, 2010). Several recent papers discuss the implications of the emergence of these technologies for BPM (e.g., Beverungen et al., 2017; Mendling et al., 2017; Oberländer, Röglinger, Rosemann, & Kees, 2017). These technologies will likely affect how organizations design and execute business processes in the future. However, it is not clear in which specific way they will change BPM.

This paper summarizes the research background and the major arguments of a panel discussion at the 15th International Conference on Business Process Management. The panel discussed to what extent the emergence of recent technologies including machine learning, robotic process automation, and blockchain will reduce the human factor in business process management. As Shazia Sadiq highlighted, these technologies have a broad potential to affect BPM; however, it is not clear whether this impact will yield a peaceful decentralization (Star Trek scenario) or of darkness and extinction (Terminator scenario). Thus, this paper also contributes to our understanding of what impact these emerging technologies will have on the way processes are designed.

The paper proceeds as follows. In Section 2, we overview BPM and summarize research that discusses the impact of technology on business processes. In Section 3, we sketch some of the emerging technologies and investigate their impact at the task level and the coordination level of business processes. In Section 4, we discuss challenges and opportunities for research. We provide an edited transcript of the panel discussed in Appendix A.

2 Business Process Management and Technological Impact

In this section, we overview BPM with the BPM lifecycle’s assistance. New technologies allow one to design processes in novel ways. With reference to the redesign phase of this lifecycle, we discuss how technology affects the way how one can improve processes.

2.1 BPM Lifecycle

The BPM lifecycle model describes how the different management activities associated with BPM relate to one another. At the single process level, the lifecycle has five different phases: process discovery, process analysis, process redesign, process implementation, and process monitoring (see Figure 1) (Dumas et al., 2013). At its heart, the model illustrates how one can organize a BPM project or a BPM initiative such that it arrives at an improved process.

Volume 43 10.17705/1CAIS.043XX Paper XXX

Communications of the Association for Information Systems 3

Discovery

Implementation

The BPM lifecycle starts with the process discovery phase. It focuses on one specific process. This phase focuses on producing detailed descriptions of a business process as it currently exists (i.e., the “so-called” or “as-is” process model). During process analysis, one applies analytical tools and techniques in order to determine a business process’s current weaknesses. Process redesign addresses the most important weaknesses and yields a reworked design of the process (i.e., a “to-be” process model). One subsequently uses this model as the basis for process implementation. Process implementation refers to the various steps to put the to-be process into operation, such as implementing information systems and measures to facilitate organizational change. In the process-monitoring phase after one has implemented the redesigned process, one continuously collects and analyzes execution data for their compliance with performance and conformance objectives. Failing to meet objectives or changes of the goals and the business environment can trigger new iterations of the BPM lifecycle.

Subjecting a business process to the management activities of the BPM lifecycle can lead to improvements at the task and coordination levels. An organization achieves improvements at the task level when it improves the duration, the costs, the quality, or the flexibility of a singular task. An organization achieves improvements at the coordination level when the overall organization of handoffs between the tasks leads to faster processing, lower costs, better quality, or more flexibility. Some indications suggest that striving for improvements at the coordination level might have a relatively stronger impact on process performance than improving singular tasks. Blackburn (1992) investigated the flow-time efficiency of business processes in various industries and found that the cycle time of most business processes contains more than 95 percent of waiting time. At least for speeding up a business process, this finding means that reducing the waiting time between the tasks (coordination level) is more likely to improve flow-time efficiency than reducing the processing time of individual tasks (task level). One needs to keep this finding in mind when we discuss the impact of specific technologies on a business process: the technology might have a dominant impact at the task or the coordination level.

2.2 Technological Impact on Business Processes

New technologies affect how organizations execute and coordinate tasks in a process. Thus, one can see a new technology’s impact most visibly in the redesign phase of the BPM lifecycle and, in particular, in specific redesign heuristics. Reijers and Mansar (2005) present an extensive list of such heuristics. Many of these heuristics explicitly refer to information technology as a means to achieve process improvements. For instance, the task automation heuristic suggests that one should take an existing task and subject it to automation. This heuristic relates to the task level. This heuristic ideally produces a faster, cheaper, and more accurate execution of the task. The interfacing heuristic represents another example. It incorporates the idea that organizations can use standardized interfaces to integrate their operations with partners’ and customers’ information systems in order to make processes faster and more reliable. This heuristic impacts the coordination level more strongly than the task level. These heuristics describe two examples of information technology that affect business processes.

The 1990s saw a strong wave of business process reengineering (Hammer & Champy 1993) together with major investments in information technology newly introduced to the market back then. At the same time, researchers, including Brynjolfsson (1993), observed a productivity paradox of information technology.

Volume 43 10.17705/1CAIS.043XX Paper XXX

4 How do Machine Learning, Robotic Process Automation and Blockchains Affect the Human Factor in BPM?

Apparently, investments in information technology did not always lead to productivity gains. Some of the works that have tried to resolve this paradox demonstrate that productivity gains from information technology investments require organizations to change their business processes in order to reap the potential benefits (Mukhopadhyay, Rajiv, & Srinivasan, 1997; Grover, Teng, Segars, & Fiedler 1998). From the perspective of new technology, Mooney, Gurbaxani, and Kraemer (1996) distinguish automational effects, informational effects, and transformational effects. Automational effects emerge when an organization uses a new technology to automate tasks that it previously did manually or with partial system support. Informational effects materialize from better tracking, monitoring, and analytical insights. Transformational effects relate to the changes in the mechanisms of coordination, which include disintermediation, outsourcing, or offshoring.

3 Emerging Technologies

In this section, we focus on three specific technologies that might affect business processes’ potential to automate tasks and facilitate new ways of coordination: machine learning, robotic process automation, and blockchains. We briefly sketch their central characteristics and point to more detailed references.

3.1 Machine Learning

Machine learning is a branch of the artificial intelligence research area. One prominent category of machine-learning applications is classification (Bishop, 2006). One can find classification tasks in various domains that require expert judgment, such as in healthcare (e.g., determining if someone has a tumor), law (e.g., determining whether to sentence someone for a crime), or construction (e.g., determining whether a certain construction would be stable). With the availability of big data in certain application domains, the potential of applying machine learning for classification has also increased. For instance, Sim (2016) emphasizes the importance of technologies such as IBM Watson for various diagnostic tasks in a medical context. One also needs big data to train machine-learning techniques such that they can provide accurate classification results. In this way, machine learning has the potential to partially automate a broader spectrum of tasks that experts have conducted in the past. It might also help to coordinate different tasks in a business process. In the context of BPM, these observations raise the questions for which specific application scenarios machine learning can be effectively devised and which type of training data is required to make it useful in a practical setting.

3.2 Robotic Process Automation

Robotic process automation (RPA) is an industrial response to the huge amount of manual work that individuals perform on a daily, weekly, or monthly basis to support a broad array of high-volume business processing (Aguirre & Rodriguez, 2017, Lacity & Willcocks, 2016). RPA is mostly associated with the task level. The application areas include finance and accounting, IT infrastructure maintenance, and front-office processing. The so-called robots are software programs that interact with systems such as enterprise resource planning and customer relationship management systems. The robots can gather data from systems and update them by imitating manual screen-based manipulations. From a business perspective, RPA solutions are appealing because they automate repetitive tasks while being minimally invasive into the overall processing that they support. An increasing number of organizations have begun to adopt RPA solutions recently; however, this growth might diminish in the future when the next generation of enterprise resource planning systems and IT infrastructure directly incorporates services for accessing data and making updates. RPA raises interesting academic research questions such as how to design and program robots and to integrate them with BPM systems, how to leverage RPA as a vehicle to support AI-enhanced processes, and how to use artificial intelligence techniques to program RPA solutions based on goals.

3.3 Blockchain

Blockchain, the technology underlying crypto-currencies such as Bitcoin, is a distributed ledger technology that enables organizations to engage in transactions without the need for a commonly trusted authority. It is a promising technology at the coordination level and a potential infrastructure for facilitating inter-organizational business processes. Its key strength is that it supports transactions between parties that do not trust each other over a computer network in which trust emerges from a combination of peer-to-peer technologies, consensus making, cryptography, and market mechanisms. Smart contracts are user-definable programs that the network of computer nodes in a blockchain executes. With the addition of these smart contracts, one can design the control logic between transactions in order to meet a diverse set of use

Volume 43 10.17705/1CAIS.043XX Paper XXX

Communications of the Association for Information Systems 5

cases that span the financial industry, logistics and supply chains, healthcare, sharing economy, and many more. Proposals from the BPM research community include using smart contracts to express processes, particularly inter-organizational ones, in an imperative form such as BPMN (Weber et al., 2016), in an artifact-centric form (Hull et al., 2016), or in a rule form (Mery & Selman, 2017). In this way, large parts of the business logic of inter-organizational business processes can be compiled from process models into smart contracts to ensure that the joint process is correctly executed. Executing inter-organizational business processes using smart contracts on a blockchain can remove several barriers (Weber et al., 2016, Mendling et al., 2018). First, the blockchain can serve as an immutable public ledger, so that participants can review a trustworthy history of messages. Second, smart contracts can offer independent process monitoring from a global viewpoint. Third, encryption can ensure that data relevant for making decisions is visible while the remaining data is only visible to the process participants that require it. Blockchain technology raises interesting research questions such as how to devise novel execution and monitoring systems for inter-organizational business processes, how to define appropriate mechanisms for process evolution and adaptation, and how to identify patterns of redesigning processes using blockchain technology.

4 Impact of Emerging Technologies on Business Process Management

In this section, we discuss the impact of the various emerging technologies on BPM. We first discuss the impact at the task level and then at the coordination level.

4.1 Impact at the Task Level

To determine the impact of the various technologies, we follow Autor (2015) and distinguish three different types of tasks:

1. Routine tasks are explicit and codifiable. They include the calculations involved in bookkeeping; the retrieving, sorting, and storing of structured information in association with clerical work; and the precise execution of repetitive physical operations in a stable environment.

2. Abstract tasks require problem-solving capabilities, intuition, creativity, and persuasion. Tasks of this kind are typically associated with professional, technical, and managerial occupations. They require employees with a high degree of education and analytical capabilities. They emphasize inductive reasoning, communication, and professional expertise in open and underspecified contexts.

3. Manual tasks require situational adaptability, visual and language skills, and personal interactions. Manual tasks typically characterize food preparation and service jobs, cleaning and janitorial work, grounds cleaning and maintenance, health assistance, and jobs in security and protection services.

Emerging technologies will likely strongly impact routine tasks since they often provide the potential to benefit from what Mooney et al. (1996) call automational effects. As such, these technologies could displace workers in these routine tasks because they follow precise, well-understood procedures that can be either codified or mimicked. Machine learning and RPA will likely contribute to this trend. The panel discussion highlighted data entry and data validation as examples of routine tasks increasingly replaced by automatic solutions that companies such as Parlamind provide. As a consequence, we might observe a substantial decline in employment in clerical and administrative support.

Emerging technologies could also have a strong impact on abstract tasks. For these tasks, we might see informational effects. The panel discussion emphasized that systems already yield much better results for tasks such as diagnosing skin cancer. A serious challenge for these tasks involves the trust in the correctness and accuracy of the solutions. The panel described one example: the process of assessing whether to grant an asylum application or not. German authorities trialed a prototype system that could have sped up the application processing drastically; however, the country did not end up using it because decision makers lacked trust in its accuracy.

Emerging technologies can also have a transformational effect on manual tasks. Applications related to the Internet of things, Industry 4.0, and the industrial Internet contribute to these developments. For example, in a classical picking process in a warehouse, workers pick the products from their respective positions. Companies such as Amazon have introduced picking robots that connect to the order information systems.

Volume 43 10.17705/1CAIS.043XX Paper XXX

6 How do Machine Learning, Robotic Process Automation and Blockchains Affect the Human Factor in BPM?

In this way, the company has transformed the picking process from manual work to work that machines perform.

One generally finds jobs intensive in either abstract or manual tasks at opposite ends of the occupational skill spectrum: professional, managerial, and technical jobs are on the one end and service and laborer jobs on the other. The computerization of routine tasks likely leads to the simultaneous growth of high-education, high-wage jobs at one end and low-education, low-wage jobs at the other end. Both developments will take place at the expense of middle-wage, middle-education jobs—a phenomenon that Goos and Manning (2003) call “job polarization”. Various economic studies at different levels of abstraction have also confirmed this phenomenon (Frey & Osbourne, 2017). However, the panel emphasized that jobs with routine tasks will continue to exist because the emerging technologies still have too many limitations. Currently, they are not profitable for tasks not highly standardized.

4.2 Impact at the Coordination Level and Work Organization

While emerging technologies may have a substantial impact on separate tasks, one needs to remember that any job involves more than one task. Many middle-wage, middle-education jobs include routine tasks but not exclusively so. The automation of routine tasks generally enhance the more complex tasks that such jobs comprise and that automation cannot replace (Autor, 2015). One can see as much in particular at a level where one needs to coordinate the tasks of many parties. Most business processes draw from a variety of inputs: labor, capital, intellect, creativity, technical skills, intuition, rules, and so on. Typically, each of these inputs plays an essential role. Thus, improving one task does not make another superfluous. In other words, productivity improvements in one set of tasks will likely increase the economic value of the remaining tasks either in a single job or a process as a whole.

One can find an iconic representation of this idea in the O-ring production function that Kremer (1993) has studied. In the O-ring model, failure of any one step in the production chain leads the entire production process to fail. Conversely, improvements in the reliability of any given link increase the value of improvements in all of the others. Intuitively, if n  1 links in the chain are reasonably likely to fail, the fact that link n is somewhat unreliable has little consequence. If the other n  1 links are made reliable, then the value of making link n more reliable rises as well. Analogously, when automation or computerization makes some steps in a work process more reliable, cheaper, or faster, the value of the remaining human links in the production chain also increases. Benefits in this dimension might result from easier coordination of inter-organizational business processes using blockchains. The panel discussed the case of AgriDigital that achieves such improvements in the agricultural sector.

Kremer (1993) discusses the application of the O-ring model for the case of automatic teller machines (ATMs). ATMs appeared in the 1970s and their number in the U.S. economy quadrupled from approximately 100,000 to 400,000 between 1995 and 2010. One might expect that such machines would have wiped out the job of bank tellers in that period. Yet, U.S. bank teller employment actually rose, albeit modestly, from 500,000 to approximately 550,000 over the 30-year period from 1980 to 2010 (although, given the growth in the labor force in this time interval, these numbers do imply that bank tellers declined as a share of overall U.S. employment). Bessen (2015) explains this somewhat paradoxical development in observing that two forces worked in opposite directions. First, by reducing the cost of operating a bank branch, ATMs indirectly increased the demand for tellers: the number of tellers per branch fell by more than a third between 1988 and 2004, but the number of urban bank branches rose by more than 40 percent. Second, as the routine cash-handling tasks of bank tellers receded, information technology also enabled a broader range of bank personnel to become involved in customer service. Increasingly, banks recognized the value of tellers supported by information technology as salespersons who forge relationships with customers and introduce them to additional bank services such as credit cards, loans, and investment products.

5 Impact beyond Singular Business Processes

Clearly, the emerging technologies we mention in this paper impact more than singular processes. The panel discussed the things they might impact and the challenges that this impact might bring for society. We summarize this discussion in seven points.

1. : the panelists expect that a good share of today’s job profiles will change or disappear in the next decade. Frey and Osbourne’s (2017) model, for instance, sees a high probability of computerization for jobs such as dishwashers, court clerks, and telemarketers, which could imply that people have to become more flexible and change jobs more often than in

Volume 43 10.17705/1CAIS.043XX Paper XXX

Communications of the Association for Information Systems 7

the past. At least a share of the workforce will find having to become more flexible challenging. However, it does not mean that our society will run out of work. The past two centuries of automation and technological progress have not made human labor obsolete. The employment-to-population ratio rose during the 20th century, and, although the unemployment rate fluctuates cyclically, we have not seen any apparent long-run increase in unemployment according to Autor (2015). Two effects compete here: technology’s destructive effect of labor substitution and a capitalization effect of rising employment in sectors that achieve productivity gains (Frey & Osbourne, 2017). It is difficult to foresee how these effects will balance out.

2. : the panelists observed that the emerging technologies mentioned often have low acceptance. As for why, one reason may concern the level of perceived behavioral control (Venkatesh, Morris, Davis, & Davis, 2003). Indeed, technologies such as machine learning, RPA, and blockchain are complex and difficult to understand, which might explain low perceived behavioral control. Paradoxically, trust in human experts is high even though they often do not agree or come to consistent diagnoses (Schön 1983). In particular, solutions based on artificial intelligence need techniques that explain automatic decisions. Otherwise, people and decision makers may not adopt fast enough even though the technology has high factual accuracy.

: new technologies have effects that can be judged as good and bad from an ethical perspective. On the downside, artificial intelligence-based solutions might simply adopt the biases and prejudices that the training data includes. Such biases are concerns of ethical standards in systems engineering (Spiekermann, 2015). On the bright side, technologies have the potential to speed up processes that people find stressful due to their long duration. The panel featured the example of a partially automated asylum application-handling process. Beyond that, technology has also the potential to make business processes fairer and less susceptible to corruption.

4. : the panelists observed that organizations often use emerging technologies mentioned to improve customer experience. New process designs increasingly use insights from design thinking (Norman, 2013). Technologies such as chatbots offer scalable solutions for customer communication and interaction, which were formerly too expensive with human workers. However, in this context, one faces challenges in balancing automatic interaction and human interaction. Customers might or might not realize that chatbots serve them. In case they realize, one needs to question how they will act and perceive the interaction.

5. : one can make similar observations about the design of the workplace and the support of office workers. Research has established that job design has an impact not only on performance but also on creativity and employee wellbeing (Oldham & Fried 2016). Using emerging technologies can contribute to building an attractive workplace. In specific scenarios, such technologies might also have the potential to protect workers from risks (e.g., sending a remotely navigated drone instead of humans to a contaminated area). An important question concerns how one can best integrate automated tasks and human work.

6. : the panelists observed that novel information technologies have the potential to make people happier and more satisfied with their life. For instance, Ibarra et al. (2016) describe how tools can help older people to make online contributions. The panel also mentioned the case of elderly people using online tools to make appointments for knitting together and the case of education management systems adapting to the pace of the learner. On the downside, various actors have increasingly begun to use social media to manipulate elections and to disintegrate society. Currently, we have no clear account on the balance of benefits and drawbacks of these technologies from a social perspective.

7. : the panel highlighted that regulations are often discussed as a means to handle the impact of emerging technologies such as blockchains and cryptocurrencies. Blind (2016) high¬lights that regulations have an ambivalent impact: empirical evidence shows the dampening effect of compliance cost and stimulating effects of regulatory incentives. The panelists mentioned the healthcare sector as an example where regulations hinder the adoption of new technologies. On the other hand, anecdotal evidence suggests that entrepreneurs in the blockchain space value regulatory clarity because it gives them certainty regarding the legality and taxation of their ventures. In many cases, national legislators and regulators and supranational organizations have or will become active in setting the rules regulating the usage of specific new technologies.

Volume 43 10.17705/1CAIS.043XX Paper XXX

8 How do Machine Learning, Robotic Process Automation and Blockchains Affect the Human Factor in BPM?

All these seven aspects require the research efforts of interdisciplinary teams. Insights from computer science, psychology, business administration, economics, engineering, political sciences, law, and other studies have to be integrated to investigate them in an adequate way. Also, curricula will have to evolve beyond the narrow boundaries of specialized fields in order to develop a broader perspective on these developments. Business processes will continue to be relevant research subjects in understanding the impact of new information technology on the profitability of existing business models and the emergence of new ones. We call for the BPM research community to reach out to these neighboring disciplines to study the impact of emerging technologies such as RPA and blockchains and directions for further improving them.

Acknowledgments

We thank the organizers of the 15th International Conference on Business Process Management, in particular Josep Carmona, for the opportunity to run this panel. We also thank the Tutorial and Panel Chairs of the conference Joaquin Ezpeleta (UZ, Zaragoza), Dirk Fahland (TU/e, Eindhoven) and Barbara Weber (DTU, Copenhagen) for their encouragement to organize this panel. Finally, we thank the audience for their active participation in the discussion. Unfortunately, we did not know the names of all persons who asked questions. We refer to these individuals as “person in the audience” in the transcript. Finally, we thank Adam LeBrocq for his great help in improving the stylistic quality of this paper.

Volume 43 10.17705/1CAIS.043XX Paper XXX

Communications of the Association for Information Systems 9

References

Aguirre, S., & Rodriguez, A. (2017). Automation of a business process using robotic process automation (RPA): A case study. In J. Figueroa-García, E. López-Santana, J. Villa-Ramírez, & R. Ferro-Escobar (Eds.), Applied computer sciences in engineering (pp. 65-71). Berlin: Springer.

Atzori, L., Iera, A., & Morabito, G. (2010). The Internet of things: A survey. Computer Networks, 54(15), 2787-2805.

Autor, D. H. (2015). Why are there still so many jobs? The history and future of workplace automation. The Journal of Economic Perspectives, 29(3), 3-30.

Bessen, J. (2015). Toil and technology. Finance and Development, 52(1), 16-19.

Beverungen, D., Müller, O., Matzner, M., Mendling, J., & vom Brocke, J. (2017). Conceptualizing smart service systems. Electronic Markets.

Bishop, C. M. (2006). Pattern recognition and machine learning. Berlin: Springer.

Blackburn, J. D. (1992). Time-based competition: White-collar activities. Business Horizons, 35(4), 96-101.

Blind, K. (2016). The impact of regulation on innovation. In J. Edler, P. Cunningham, A. Gok, & P. Shapira (Eds.), Handbook of innovation policy impact (pp. 450-482) Cheltenham, UK: Edward Elgar Publishing.

Brynjolfsson, E. (1993). The productivity paradox of information technology. Communications of the ACM, 36(12), 66-77.

Daily, M., Medasani, S., Behringer, R., & Trivedi, M. (2017). Self-driving cars. Computer, 50(12), 18-23.

Dumas, M., La Rosa, M., Mendling, J., & Reijers, H. A. (2013). Fundamentals of business process management. Heidelberg: Springer.

Dumas, M., van der Aalst, W. M., & ter Hofstede, A. H. (2005). Process-aware information systems: Bridging people and software through process technology. New York: Wiley.

Frey, C. B., & Osborne, M. A. (2017). The future of employment: How susceptible are jobs to computerisation? Technological Forecasting and Social Change, 114, 254-280.

Goos, M., & Manning, A. (2007). Lousy and lovely jobs: The rising polarization of work in Britain. The Review of Economics and Statistics, 89(1), 118-133.

Grover, V., Teng, J., Segars, A. H., & Fiedler, K. (1998). The influence of information technology diffusion and business process change on perceived productivity: The IS executive's perspective. Information & Management, 34(3), 141-159.

Hammer, M., & Champy, J. (1993). Reengineering the corporation. New York: Harper Collins.

Heilprin, L. (1964). American Documentation Institute Committee on organization of information: Report for 1962-63. Journal of the American Society for Information Science, 15(4), 274–288.

Hirschheim, R. A. (1985). Office automation: A social and organizational perspective. Chichester, UK: Wiley.

Hull, R., Batra, V. S., Chen, Y.-M., Deutsch, A., Heath, F. F. T., & Vianu, V. (2016). Towards a shared ledger business collaboration language based on data-aware processes. In Q. Z. Sheng, E. Stroulia, S. Tata, & S. Bhiri (Eds.), Proceedings of the 14th International Conference on Service-oriented computing (LNCS vol. 9936, pp. 18-36). Berlin: Springer.

Ibarra, F., Korovina, O., Baez, M., Casati, F., Marchese, M., Cernuzzi, L., & Barysheva, G. A. (2016). Tools enabling online contributions by older adults. IEEE Internet Computing, 20(5), 58-65.

Kremer, M. (1993). The O-ring theory of economic development. The Quarterly Journal of Economics, 108(3), 551-575.

Lacity, M. C., & Willcocks, L. P. (2016). Robotic process automation at Telefonica O2. MIS Quarterly Executive, 15(1), 21-35.

Mendling, J. Weber, I., van der Aalst, W., Brocke, J. V., Cabanillas, C., Daniel, F., Debois, S., Di Ciccio, C., Dumas, M., Dustdar, S., Gal, A;, García-Bañuelos, L., Governatori, G., Hull, R., La Rosa, M., Leopold,

Volume 43 10.17705/1CAIS.043XX Paper XXX

10 How do Machine Learning, Robotic Process Automation and Blockchains Affect the Human Factor in BPM?

H., Leymann, F., Recker, J., Reichert, M., Reijers, H.A., Rinderle-Ma, S., Solti, A., Rosemann, M., Schulte, S., Singh, M.P., Slaats, T., Staples, M., Weber, B., Weidlich, M., Weske, M., Xu, X., & Zhu, L. (2018). Blockchains for business process management—challenges and opportunities. ACM Transactions on Management Information Systems, 9(1), 1-16.

Mery, S., & Selman, D. (2017). Make your blockchain smart contracts smarter with business rules. IBM developerWorks. Retrieved from https://www.ibm.com/developerworks/library/mw-1708-mery-blockchain/1708-mery.htm l

Mooney, J. G., Gurbaxani, V., & Kraemer, K. L. (1996). A process oriented framework for assessing the business value of information technology. ACM SIGMIS Database, 27(2), 68-81.

Mukhopadhyay, T., Rajiv, S., & Srinivasan, K. (1997). Information technology impact on process output and quality. Management Science, 43(12), 1645-1659.

Nakamoto, S. (2008). Bitcoin: A peer-to-peer electronic cash system. Retrieved from https://bitcoin.org/bitcoin.pdf

Norman, D. (2013). The design of everyday things: Revised and expanded edition. New York, NY: Basic Books.

Oberländer, A. M., Röglinger, M., Rosemann, M., & Kees, A. (2017). Conceptualizing business-to-thing interactions–A sociomaterial perspective on the Internet of Things. European Journal of Information Systems.

Oldham, G. R., & Fried, Y. (2016). Job design research and theory: Past, present and future. Organizational Behavior and Human Decision Processes, 136, 20-35.

Reijers, H. A., & Mansar, S. L. (2005). Best practices in business process redesign: An overview and qualitative evaluation of successful redesign heuristics. Omega, 33(4), 283-306.

Schön, D. (1983). The reflective practitioner: How practitioners think in action. London: Temple Smith. Shawar, B. A., & Atwell, E. (2007). Chatbots: Are they really useful? LDV Forum, 22(1), 29-49.

Sim, I. (2016). Two ways of knowing: Big data and evidence-based medicine. Annals of Internal Medicine, 164(8), 562-563.

Spiekermann, S. (2015). Ethical IT Innovation: A value-based system design approach. Boca Raton, FL: CRC Press.

van der Aalst, W., & van Hee, K. M. (2004). Workflow management: Models, methods, and systems. Cambridge, MA: MIT Press.

Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of information technology: Toward a unified view. MIS Quarterly, 27(3), 425-478.

Weber, I., Xu, X., Riveret, R., Governatori, G., Ponomarev, A., & Mendling, J. (2016). Untrusted business process monitoring and execution using blockchain. In La Rosa, M., Loos, P., & Pastor, O. (Eds.), Proceedings of the 14th International Conference on Business Process Management (LNCS vol. 9850, pp. 329-347). Berlin: Springer.

Volume 43 10.17705/1CAIS.043XX Paper XXX

Jasa Konsultasi Disertasi Keuangan (1) dan Perbankan S3, hubungi WA 0821 2230 7021

Selasa, 09 Agustus 2022

Contoh Tesis mengenai OFDM modulation

Tidak ada komentar:

Posting Komentar