Performance Evaluation of IDMA

Scheme in Wireless Communication

A Thesis
Submitted in Partial Fulfillment of the Requirements
for the Degree of

Doctor of Philosophy
in
Electronics & Communication Engineering



by
Manoj Kumar Shukla



Department of Electronics & Communication Engineering
Motilal Nehru National Institute of Technology,
Allahabad (India)
November- 2010

Undertaking

I declare that the work presented in this thesis titled Performance Evaluation of IDMA
Scheme in Wireless Communication, submitted to the Department of Electronics &
Communication Engineering, Motilal Nehru National Institute of Technology, Allahabad, for
award of the Doctor of Philosophy in Electronics & Communication Engineering, is my original
work. I have not plagiarized or submitted the same work for the award of any other degree. In
case this undertaking is found incorrect, I accept that my degree may be unconditionally
withdrawn.


November, 2010
Allahabad (Manoj Kumar Shukla)
Certificate


Certified that the work contained in this thesis titled Performance Evaluation of IDMA Scheme in
Wireless Communication submitted by Manoj Kumar Shukla, has been carried out under our
supervision and that this work has not been submitted elsewhere for a degree.


Dr. Sudarshan Tiwari Dr. V.K. Srivastava
Professor and Head Professor
Department of Electronics & Comm. Engg. Department of Electronics & Comm. Engg.
Motilal Nehru National Institute of Technology Motilal Nehru National Institute of Technology
Allahabad, INDIA Allahabad, INDIA


Acknowledgements

I am highly indebted to my thesis supervisors, Professor Sudarshan Tiwari and Professor V.K.
Srivastava for their kind support and guidance without which this thesis would not been possible.
Their sound knowledge, enthusiasm to research, and guidance was invaluable to me since the day I started
as a research scholar. Their untiring encouragement has always been endless source of motivation for me.
I can never thank them enough. I feel to be lucky to work under their invaluable guidance during the
research work in MNNIT, Allahabad.
I am extremely grateful to Professor Rajiv Tripathi for his valuable encouragement during the entire
research work. He was really helpful especially during the stressed moments during all these years. I have
always found him a true reliever. Thank you sir!
I am also thankful to Professor T.N. Sharma, Professor H.N. Kar, and all the faculty members of the
department for maintaining a congenial research environment. I am also thankful to many colleagues
who have enhanced my understanding of the subject, in particular to Dr. P.C. Upadhyaya, Dr.
V.S. Tripathi and Dr. D.K. Kothari. These colleagues and valued friends, too numerous to be
mentioned, have influenced my views concerning various aspects of wireless communication. I
am also grateful to Rajesh Verma, V.K. Dwivedi, A. Raghuvanshi, Arun Prakash, Malaya Hota,
Prashant Shah, Subodh Waria, and many others with whom I enjoyed an association.
I also acknowledge my valuable associations with the faculty members of Department of
Electronics Engineering, Harcourt Butler Technological Institute, Kanpur, INDIA, in particular
with Prof. G.P. Bagaria, Dr. Rachna Asthana, Mrs. Rajani Bisht, Dr. Krishna Raj, Mr. Ashok
Shankhwar, Mr. Ashutosh Singh, Mr. Ram Chandra Singh Chauhan and many other valued
colleagues. My sincere thanks are also due to administration of Harcourt Butler Technological
Institute, Kanpur, INDIA, for supporting my research.
i

I feel particularly indebted to my guides for her skilful assistance in correcting the final
manuscript in MS WORD. Finally, my sincere gratitude is due to the numerous authors listed in
the Author Index as well as to those whose work was not cited owing to space limitations for
their contributions to the state of the art, without whom this thesis would not have materialized.


Manoj Kumar Shukla
Department of Electronics & Communication Engineering
Motilal Nehru National Institute of Technology, Allahabad, INDIA


















ii

Synopsis
The development of wireless cellular communication systems has evolved from the first-
generation (1G) of analogue stage to second-generation (2G) and third-generation (3G) of digital
stage. Now, due to market oriented demands, it is has stepped into fourth-generation (4G) of
broadband stage. As per recommendations of International Mobile Telecommunications-2000
(IMT-2000), the future wireless communication is bound to occupy the features including
high-speed data and broadband transmission, high capacity to support a huge number of
simultaneous users, global mobility, high security, and scalable quality of service (QoS) along
with low cost for both operators and subscribers. The above features are imposing technical
challenges on system design and stimulating various research topics on capacity, complexity and
performance. In order to increase the capacity of wireless networks, various multiple access
schemes have been reported in the literature. The credit of most competent multiple access
scheme in 2G systems goes to CDMA scheme which offers an even better bandwidth-efficiency
than TDMA and FDMA schemes, however, its implementation is quite difficult due to
involvement of rather complex technologies including complex power-control, and multiuser
detection techniques etc. The performance of CDMA scheme is mainly limited by multiple
access interference (MAI) and inter-symbol interference (ISI). In conventional CDMA
systems, the spreading sequences were employed for the purpose of user separation, however
due to poor aperiodic correlations amongst spreading codes including Gold, Walsh, Kasami
sequences, the direct-sequence (DS) spreading mechanism demonstrates low spreading
efficiency (SE) in case of high user count.
iii

The requirement of alternate mechanism for user separation has been solved by Interleave-
Division Multiple-Access (IDMA) scheme, in which, most of above stated problems do not exist
due to application of user-specific interleavers having low cross-correlation amongst them. The
interleaved data resulted from user-specific interleavers, demonstrates better orthogonality
amongst each other in the channel. The condition of orthogonality is maintained for reducing the
risk of collision amongst the interleavers during communication process.
In IDMA scheme, orthogonal interleavers are employed as the only means for user separation
and, hence, are referred as the heart of the scheme. The selection of interleaver along with
optimum design methodology for IDMA system leads to satisfactory spectral efficiency.
During initial run of IDMA scheme, random interleavers (RI) were employed for user
separation. Later, random interleavers were replaced by master random interleavers (MRI) in
order to reduce the memory requirement raised due to storage of random interleavers at
transmitter and receiver ends. Many more interleavers are reported in literature but most of
them are based on methodology of selection of user-specific interleavers amongst available
random interleavers. Still, the problem of computational complexity involved in
interleaving and de-interleaving mechanism is unresolved for user specific interleavers.
These problems in user-specific interleavers have provided the motivation for development of
an optimum interleaving mechanism for IDMA scheme.
In this thesis, an optimum interleaving mechanism has been proposed named as tree based
interleaving (TBI) mechanism, for the solutions related to computational complexity,
bandwidth requirement and optimization of memory size, at transmitter and receiver ends.
The orientation of the work has been maintained towards the analysis and design of
iv

proposed tree based interleaving (TBI) mechanism for IDMA scheme fulfilling the
requirement of orthogonality and easy implementation.
In the beginning of thesis, the mechanism of interleaving with necessary conditions is
presented. Later, the performance and analysis of proposed TBI mechanism with IDMA scheme
has been presented. Apart from the bit error rate (BER) performance analysis, the
interleavers have also been analyzed on the basis of memory requirement and
computational complexity at transmitter and receiver ends. The performance evaluation of
IDMA scheme with proposed tree based interleaving (TBI) mechanism, in uncoded and coded
environments, has been duly investigated. After investigation, it is noticed that the computational
complexity of tree based interleaver is extremely less in comparison to that for master random
interleaver while it is marginally higher to that occurring in case of random interleaver.
However, the bandwidth and memory requirement of proposed tree based interleaver is found to
be considerably less than that of random interleaver and slightly higher to that of master random
interleaver. It has been observed that BER performance of this interleaver is similar to that of
random interleaver and master random interleaver.
The second goal of the thesis is to evaluate BER performance of IDMA scheme with
maximal ratio combining (MRC) diversity, for proposed TBI mechanism along with RI and
MRI mechanisms with various architectures. After simulations, the BER performance of
proposed TBI mechanism is observed to be very near to that of random interleaver.
Further, the correlation analysis of all the concerned interleavers have been carried out in this
thesis which reveals the behavior of interleavers with increment in user count. It is observed
that the cross-correlation of TBI mechanism is almost similar to that of MRI and RI mechanisms.
v

It again confirms that BER performance of TBI mechanism should be similar to that of other
mechanisms. The increment in multiple access interference (MAI) is also observed with
increment in user count for all the interleaving mechanisms.
Finally, the proposed tree based interleaving mechanism has been implemented on field-
programmable gate-array (FPGA) system for observing the performance related to hardware
requirements and timing constraints in comparison to that required with RI and MRI
mechanisms, for IDMA systems. During the implementations, it is observed that the hardware
requirement of tree based interleaving (TBI) mechanism is at the minimum level in comparison
to other considered mechanisms. The timing constraints of TBI mechanism are also found to be
at its minimum level. In addition to it, the tree based interleaving mechanism inherits comparably
lesser hardware complexity to that required for master random interleaving mechanism due to
requirement of lesser looping operations. After the analysis, it has been observed that proposed
tree based interleaver requires least hardware along with least timing constraints for its operation.
The proposed tree based interleaving (TBI) mechanism has demonstrated optimum overall BER
and other performances in comparison to RI, and MRI mechanisms. Thus, the proposed
interleaving mechanism along with IDMA scheme provides an alternative to conventional
CDMA scheme for future wireless communication systems.






vi

Table of Contents

Acknowledgements i
Synopsis iii
Table of Contents vii
List of Figures x
List of Table xiii
Glossary xv

1. Introduction 1
1.1. Development of Wireless Communication Systems 1
1.2. Multiple Access Schemes 5
1.2.1. FDMA Scheme 5
1.2.2. TDMA Scheme 6
1.2.3. CDMA Scheme 7
1.3. Motivation 8
1.4. Problem Statement 10
1.5. Research Contributions 12
1.6. Thesis Organization 13

2. Overview of Interleave-Division Multiple-Access (IDMA) Scheme 15
2.1. Introduction 15
2.2. Interleavers in Digital Communication 15
2.3. Interleavers in IDMA Scheme 16
2.4. Mechanism of Interleaving Process 17
2.5. Interleave-Division Multiple-Access (IDMA) Scheme 19
vii

2.5.1. Comparison of CDMA and IDMA Schemes 20

2.5.2. IDMA Transmitter and Receiver 21
2.5.2.1. Basic Primary Signal Estimator (PSE) Function 23
2.5.2.2. Algorithm for Chip-By-Chip Detection 24
2.5.2.3. Decoder (DEC) Function 25
2.5.3. IDMA over Multipath Channels 26
2.6. Literature Review 31
2.7. Simulation of IDMA Scheme 39
2.8. Conclusions 41

3. Performance Evaluation of Tree Based Interleaver (TBI) in IDMA Scheme 42
3.1. Introduction 42
3.2. Motivation 43
3.3. Mechanism of Tree Based Interleaver (TBI) 44
3.4. Performance Evaluation of Tree Based Interleaver 47
3.5. TBI with Unequal Power Allocation Algorithm 61
3.5.1. Unequal Power Allocation Mechanism 61
3.5.2. Numerical Results 63
3.6. Conclusions 66

4. Performance Evaluation of Tree Based Interleaver in IDMA Scheme with
Maximal Ratio Combining (MRC) Diversity 68
4.1. Introduction 68
4.2. Diversity Mechanisms 68
4.2.1. Frequency Diversity 69
4.2.2. Time Diversity 70
viii

4.2.3. Space Diversity 71

4.2.3.1. Transmit Diversity 72
4.2.3.2. Receive Diversity 73
4.3. Combining Mechanisms 74
4.3.1. Selection Combining 74
4.3.2. Maximal Ratio Combining (MRC) 76
4.3.3. Equal Gain Combining (EGC) 78
4.4. Performance Evaluation of IDMA Scheme with MRC Diversity 78
4.4.1. IDMA Scheme with Maximal Ratio Receiver Combining (MRRC)
Diversity 78
4.4.2. IDMA Scheme with Maximal Ratio Transmitter Combining (MRTC)
Diversity 80
4.5. Simulation Results 82
4.5.1. Simulation Results of IDMA Scheme using MRRC Diversity 82
4.5.2. Simulation Results of IDMA Scheme using MRTC Diversity 90
4.6. Conclusions 94

5. Correlation Analysis and FPGA Implementation of Interleavers 96
5.1. Introduction 96
5.2. Motivation 96
5.3. Design Criteria for Interleavers in IDMA Scheme 98
5.4. Correlation in Interleavers 98
5.5. Correlation Analysis of Interleavers 99
5.6. Interleaving Mechanism in IDMA Scheme 105
5.6.1. Random Interleaving (RI) Mechanism 107
5.6.2. Master Random Interleaving (MRI) Mechanism 109
ix

5.6.3. Tree Based Interleaving (TBI) Mechanism 110

5.7. Performance Comparison of Interleavers on FPGA Implementation 112
5.7.1. Summary of Hardware 112
5.7.2. Final Register Report 113
5.7.3. Device Utilization Report 113
5.7.4. Timing Summary Report 114
5.8. Conclusions 115

6. Conclusions 118
6.1. Suggestions for Further Investigations 121

References 123

Appendix 142

List of Figures
1.1 Progress in wireless communication from 1G to 4G 3
1.2 Frequency Division Multiple Access (FDMA) 5
1.3 Time Division Multiple Access (TDMA) 6
2.1 Mechanism of data interleaving 18
2.2 CDMA scheme vs. IDMA scheme 20
2.3 IDMA transmitter and receiver structure 22
2.4 IDMA transmission in single path 23
2.5 Flowchart of decoding mechanism in the receiver of IDMA scheme 26
2.6 IDMA in multipath transmission 27
2.7 Simulation of IDMA and CDMA schemes 40
2.8 Simulation of IDMA scheme with Random Interleaver 40
3.1 Interleaving strategy for Tree Based Interleaving scheme 45
3.2 Performance of Tree based Interleaver with Random Interleaver 50
3.3 Comparison of RI, MRI, and TBI for memory requirement 52
3.4 Comparison of RI, MRI, and TBI for computational complexity at
transmitter end 54
3.5 Comparison of RI, MRI, and TBI for computational complexity at
receiver end 55
3.6 Data formats of RI, MRI, and TBI mechanisms in IDMA scheme 56
3.7 Simulation of TBI in multi-user environment 57
3.8 Uncoded IDMA scheme in AWGN and Flat Rayleigh fading environment
57
3.9 Simulation of RI in coded and uncoded IDMA scheme 58
3.10 IDMA scheme in uncoded environment for variation in user count 59
3.11 Simulation of RI in coded and uncoded IDMA scheme 59
3.12 IDMA scheme in coded environment for variation in user count 60
3.13 Simulation results for 32 users without coding with RI with various data
xi

lengths 64
3.14 Simulation result for 32 users without coding with TBI with various
data lengths 65
3.15 IDMA scheme in uncoded environment for variation in user count
with RI and TBI with unequal power allocation algorithm 65
3.16 IDMA scheme in rate convolutionally coded environment for variation
in user count with RI and TBI with unequal power allocation algorithm 66
4.1 Frequency diversity Mechanism 70
4.2 Time Diversity Mechanism 70
4.3 Transmit diversity with multiple antennas at transmitter side 72
4.4 Receive Diversity having multiple antennas at receiver side 73
4.5 Mechanism of Selection combining 75
4.6 Maximal Ratio Combining (MRC) 76
4.7 IDMA with MRRC Receiver diversity 79
4.8 Transmit diversity having two transmitter and one receiver antenna 81
4.9 Performance of IDMA system with and without MRRC diversity 83
4.10 Performance of IDMA having Random Interlever with MRRC diversity
technque at various iterations count with datalength=1024,
spreadlength=16 84
4.11 Performance comparison at different data length with Random Interleaver
with MRRC Diversity 85
4.12 Performance of IDMA using tree based interleaver with MRRC Diversity 86
4.13 Simulation of Uncoded IDMA with variation in interleaver 86
4.14 Performance of Random Interlever with 1Transmitter and 2 Receive
Antenna, MRRC diversity Technique 87
4.15 Performane of master random interleaver with 1Transmitter and
2 Receive Antenna, MRRC diversity Technique 88
4.16 Performance of Tree based Interleaver With 1Transmitter and 2 Receive
Antenna, MRRC diversity Technique 88
xii

4.17 Perfromance of IDMA scheme with FEC coding with tree based
interleaver and datalength=1024 bits 89
4.18 IDMA scheme with variation in receiver count using MRRC diversity 89
4.19 Performance of IDMA system using MRTC diversity with random
interleaver 91
4.20 Performance of IDMA at different data lengths using random interleaver
with MRTC diversity Scheme 92
4.21 Uncoded IDMA scheme with variation in datalength for RI & TBI 93
4.22 Rate Conlutionaly coded IDMA scheme with variation in datalength
for RI & TBI 93
4.23 Rate Conlutionaly coded IDMA scheme with uncoded IDMA scheme 94
5.1 Mechanism for calculation of resultant user-specific cross-correlation
amongst users 101
5.2 Resultant user-specific cross-correlation for 25 users with RI, MRI,
and TBI 102
5.3 Resultant user-specific cross-correlation for 100 users with RI, MRI,
and TBI 103
5.4 Graphical view of resultant user-specific cross-correlation with RI, MRI,
and TBI 105
5.5 Block Diagram of Random Interleaving Mechanism 108
5.6 Block Diagram of Master Random Interleaver (MRI) 110
5.7 Block Diagram of Tree based Interleaving mechanism 111



xiii

List of Tables

3.1 Comparison of Memory requirement of user-specific interleavers in
IDMA scheme 51
3.2 Comparison of Computational complexity of user-specific interleavers/
deinterleavers at transmitter end 52
3.3 Comparison of Computational complexity of user-specific interleavers/
deinterleavers at receiver end 54
5.1 Peak Resultant User-Specific Cross-Correlation of RI, MRI, and TBI 104
5.2 Summary of hardware 112
5.3 Final Register report 113
5.4 Device utilization report 114
5.5 Timing summary of interleavers 115











xiv

Glossary


3GPP Third Generation Partnership Project
AMPS Advanced Mobile Phone Service
APP A Posteriori Probability
AWGN Additive White Gaussian Noise
BER Bit Error Rate
BPSK Binary Phase Shift Keying
BS Base Station
CBC Chip by Chip
CDMA Code Division Multiple Access
DEC Decoder
DL Downlink
DS Direct Sequence
EDGE Enhanced Data Rate for Global Evolution
ENC Encoder
ETACS Extended Total Access Communication System
EV-DO Evolution-Data Optimized or Evolution-Data only
FDD Frequency-Division Duplex
FDMA Frequency Division Multiple Access
xv

FEC Forward Error Correction

FPGA Field-Programmable Gate-Array
GPRS General Packet Radio Service
GSM Global System for Mobile Communications
HSDPA High-Speed Downlink Packet Access
HSPA High Speed Packet Access
IDMA Interleave Division Multiple Access
IS-136 Interim Standard-136
IS-95 Interim Standard-95
ISI Inter Symbol Interference
ITU International Telecommunication Union
LAN Local Area Network
LDPC Low Density Parity Check
LLR Log Likelihood Ratio
LTE Long Term Evolution
MAC Multiple Access Channel
MAI Multiple Access Interference
MC Multi Carrier
MC-CDMA Multi-Carrier CDMA
MIMO Multiple Input Multiple Output
MRC Maximal Ratio Combining
xvi

MRI Master Random Interleaver

MRRC Maximal Ratio Receiver Combining
MRTC Maximal Ratio Transmitter Combining
MS Mobile Station
MUD Multi User Detection
NMT Nordic Mobile Telephony
OFDM Orthogonal Frequency Division Multiplexing
PDC Personal Digital Cellular
PEG Progressive Edge Growth
PI Power Interleaver
PN Sequences Pseudo Noise Sequences
PSE Primary Signal Estimator
QoS Quality of Service
QPSK Quadrature Phase Shift Keying
RI Random Interleaver
RTT Radio Transmission Technology
SAS Singe Antenna System
SE Spreading Efficiency
SNR Signal to Noise Ratio
TACS Total Access Communication System
TBI Tree Based Interleaver
xvii

xviii

TDD Time Division Duplex

TDMA Time Division Multiple Access
TD-SCDMA Time-Division synchronous CDMA
UL Uplink
UMTS Universal Mobile Telecommunications System
UWB Ultra Wide Band
VHDL Very High Speed IC Hardware Description Language
WCDMA Wideband Code Division Multiple Access
CHAPTER 1
Introduction

Since the beginning of the 20
th
century, technologies have placed its marks with stone
line for providing new techniques and products for wireless communication.
Especially in the past three decades, wireless communication services have penetrated
into our society with an explosive growth rate.
Cellular radio was originally developed for offering phone services to mobile
subscribers. Now-a-days, it is engaged in even providing a variety of services,
including video conferencing, music or movie appreciation, games, internet access.
The demands and applications from subscribers stimulate the market and drive the
technology for further growth. On the other hand, research and development of
communication engineering are undergoing a revolution due to rapid advances in
technology.

1.1. Development of Wireless Communication Systems
Wireless cellular communication systems have evolved with the first-generation
(1G) of analogue stage using frequency division multiple access (FDMA) scheme.
Later, due to rapid advances in technologies based on market demand, it has led to
the second-generation (2G) of digital stage with time division multiple access
(TDMA) and code division multiple access (CDMA) schemes, and now it has stepped
into the third-generation (3G) with eye on fourth-generation (4G) of broadband
stage [8, 9, 12].
1

The most popular technology related to 1G is the Advanced Mobile Phone
Service (AMPS) developed in the United States (U.S.) by AT&T in the late 1970s,
and later, it was implemented by Ameritech at the end of 1983. The further
developed version of AMPS was known as Extended Total Access Communication
System (ETACS) developed in Europe in 1985. The both of these systems were
employing frequency-division duplex (FDD) and frequency-division multiple-access
(FDMA) scheme. Due to problem of slower data rate and lower user base, these
analogue systems, soon, were replaced by 2G digital systems based on time-division
multiple-access (TDMA) scheme. The representatives of 2G systems i.e. Interim
Standard-95 (IS-95) system and the Global System for Mobile Communication
(GSM), have been widely deployed throughout the world. The IS-95 system,
developed in United States of America, is mainly based on code-division multiple-
access (CDMA) scheme while GSM system, developed in Europe, is mainly based
on time-division multiple-access (TDMA) scheme.
Due to lots of technological changes and market oriented demands, mobile
communication technology has entered in 3G stage. The distinctive features of 3G
systems in comparison to 2G systems are inherited with technology of packet-
switched high-rate data transmission along with voice services. Specifically,
CDMA2000, a representative of 3G systems, builds on the packet-switched
technology along with increased data transmission rate, and backward compatibility
with original CDMA standards. It is employed primarily in North America and some
parts of Asia. Another qualified 3G candidate, wideband CDMA (WCDMA) is
referred as an evolution of the GSM technology, including aspects of TDMA and
CDMA2000 for global accessibility. The time-division synchronous CDMA (TD-
SCDMA) is mainly developed by the Datang Group, China, building on the original
2

CDMA standard to deliver multimedia data, considering its largest user base in its
own country. Figure 1.1 demonstrates the development of progress of wireless
communication from 1G to 4G, in terms of data rates, with all the technological
development observed in meantime. The CDMA and its other extended versions
such as IDMA, lie between technological developments from 2G to 3G during the
span from 1993 to 2007.

Figure 1.1: Progress in wireless communication from 1G to 4G [148]
Although the standards for further generation systems are still in formative
stages, leading companies in the industry have started some groundwork with their
researchers. Now-a-days, the techniques related to future wireless communication
have become hot topics for research all over the world.
The 3G and beyond systems have been developed to serve people's daily work
and life, and to satisfy their demands. The ultimate user needs reliable, cheaper,
secure, and low-delay voice & data services anytime and anywhere. The additional
features of the future wireless communications include high-speed data and
3

broadband transmission for huge user base, along with global mobility, & scalable
quality of service (QoS) for both operators and subscribers.
The above features are imposing technical challenges on system design and
stimulate various researches to work on topics related to capacity, complexity and
performance of the communication systems [23, 56, 20]. There are also other research
topics highly related to the physical layer in wireless systems including optimum
channel coding, detection, and diversity mechanisms. For immediate solutions of
above problems, near-capacity-achieving forward error correction (FEC) codes are
developed to enhance power efficiency while improved detection algorithms are
designed to enhance the reliability or bit-error rate (BER) performance. Diversity
techniques have been proposed to increase spectral efficiency and diversity for
accommodating more users and mitigating fading [20]. The new horizons on above
discussed topics have emerged as hot cake for researchers all over the world.
In India, cellular industry came into existence nearly in mid 1990s and since
then the average growth rate per annum has been about 85 percent [1]. By the end of
2002, the total number of cellular subscribers, in India, had increased to about 10
million subscribers. In addition to it, telecom customers have also been doubled
during last two years from 300 million to 600 million. According to Reuters India
[3], total mobile users in India, now, stand at 584.32 million, data from the sector
regulator showed, behind only China that had 777 million at the end of March 2010.
There is tremendous scope for researches in the area of wireless
communication for improving the back-bone of communication systems as per the
recommendations of International Telecommunication Union (ITU) [5]. One of the
hot-cake areas for research is improvement in technology related to multiple access
4

techniques with communication channels. In the next, section, various multiple access
techniques including recently evolved IDMA scheme will be discussed, in brief.

1.2. Multiple Access Schemes
Generally, in wireless communication, large numbers of users are involved in the
conversation at a time with each other leading to share the same wireless channel.
For sharing of wireless channel, there exist three widely deployed multiple access
schemes [8] popularly known as, frequency-division multiple-access (FDMA), time-
division multiple-access (TDMA), and code-division multiple-access (CDMA). It
will also be worth mentioning that recently, a new variant of CDMA scheme i.e.
interleave-division multiple-access (IDMA) scheme has been proposed [76].
In the next subsection, the prime multiple access schemes are being
presented in brief so as highlight their merits and demerits.

1.2.1. FDMA Scheme:
Frequency Division Multiple Access (FDMA) scheme is reffered as the most common
technique employed in analog communication systems utilizing division of entire
frequency spectrum into multiple frequencies slots to be assigned to indivdual users,
as shown in figure 1.2. With FDMA scheme, each subscriber at any given time is
assigned with particular frequency channels for transmission and reception
independently. The channel, therefore, is closed to other conversations until the
initial call is completed, or handed-off to a different channel. FDMA scheme has
been used for first generation analog communication systems.
5


Figure 1.2: Frequency Division Multiple Access (FDMA)
The scheme is referred to be inefficient due to underutilization of bandwidth.
In addition to it, FDMA systems are bound to employ a guard-band between adjacent
channels, for avoiding random Doppler shift, occurring due to the user's random
mobility. The guard-bands also reduce the probability of adjacent channels
interference, while decrease the spectral efficiency [9].

1.2.2. TDMA Scheme:

Time Division Multiple Access (TDMA) scheme improves spectrum capacity by
splitting each time period into multiple time slots. It allows each user to access the
entire radio frequency channel for the allotted time slot of during a call as presented in
figure 1.3. Other users are also allowed to share the same frequency channel at
different time slots. TDMA scheme is the dominant technology for the second
generation (2G) mobile cellular networks.

Figure 1.3: Time Division Multiple Access (TDMA)

However, the TDMA systems have to be carefully synchronized during
communication for all the users to ensure that they are received in the correct time-
6

slots and do not cause the interference to other users [7]. Since it cannot be perfectly
controlled in a mobile environment, a guard-time is inserted between each time-slot,
which reduces the probability that users will interfere, but decreases the spectral
efficiency. Also in case of bursty traffic, user has to wait for his next allotted time-
slot which ultimately slows down the data rate during communication [9].

1.2.3. CDMA Scheme:

Code Division Multiple Access (CDMA) scheme is, basically, based on spread
spectrum technology. It increases spectrum capacity by allowing all users to occupy
all channels at the same time [10]. The data streames related to indivual users are
spreaded over the whole radio band, and each voice or data call is assigned a unique
code to differentiate from the other calls carried over the same spectrum [14]. The
asynchronous CDMA system offers a key advantage in the flexible allocation of
communication resources. It is ideally suited to a mobile network where large
numbers of transmitters generating a relatively small amount of traffic at irregular
intervals individually.
The performance of CDMA scheme is mainly limited by multiple access
interference (MAI) and intersymbol interference (ISI). Its processing gain is
reduced considerably with increment in users per sector. The processing gain is
referred as a figure of merit in spread spectrum communication. Also, in CDMA
scheme, the complexity of decoder increases with increment in user count [6].
From the viewpoint of sharing communication resources, TDMA scheme is
termed to be more efficient to FDMA scheme due to better spectral effinciency
while considering the matter of message delay, FDMA scheme outperforms the
TDMA scheme. In addition to it, as quoted in [5] regarding future requirements in
7

wireless communication, it is recommended that a communication system must
possess the essential parameters for consumers including low receiver cost, de-
centralized control, diversity against fading, power efficiency, multi-media
services, high user number, and high throughput along with high spectral efficiency
[6].
For further way of efficient communication, employing spread spectrum
technology, the question regarding alternate way for user separation arises. Recently, a
new variant of CDMA scheme known as interleave-division multiple-access (IDMA)
scheme has evolved on the horizon of wireless communication [76, 77]. The IDMA
scheme employs the interleavers as the only means of user separation in order to
ensure privacy related to data of users.

1.3. Motivation
The most commonly employed multiple scheme in the world, i.e. CDMA scheme
offers an even better bandwidth-efficiency than TDMA and FDMA schemes, and has
been widely adopted in the 3G mobile cellular systems, including CDMA2000,
WCDMA, TD-SCDMA systems. It offers robust performance due to its unique
feature of processing gain. However, its successful operation is based on rather
complex technologies including complex power-control, and multiuser detection
techniques, and thus it is comparatively difficult to get implemented, when compared
with FDMA and TDMA schemes.
The CDMA mechanism is reported to be unsuitable to support QoS sensitive
multimedia traffic. It is extremely difficult to adjust data rate on-a-fly and even the
small change in data rate may result with change in processing gain [7], which further
compels to adjust transmitter power. Hence, ultimately rate change for ONE user
8

affects whole cell-wise code-assignment plan [10]. Looking into implementation
complexity of CDMA systems, the requirement of very precise power control,
powerful multi-user detection, requirement of RAKE receiver, and costly sectorized
antennas, becomes highly desirable.
In addition to it, CDMA system needs long frames for signal detection.
Therefore, it is well suited for slow-speed continuous-time transmission specially.
Apart from it, it inherits poor orthogonality of user-specific spreading codes and
merely periodic correlation functions are considered in code design process resulting
in poor aperiodic correlations amongst spreading codes for higher user count.
Therefore, a big room is left for the researchers, leading to improvement in spreading
efficiency of CDMA systems. Apart from it, only unitary codes, i.e., Gold, Walsh,
Kasami, etc. have been used in CDMA scheme.
All of the above stated problems come from the same root i.e. inefficient
Unitary codes i.e. one-code-per-user basis, used for user separation in CDMA
systems. Though, the spreading PN-sequences used for user separation in CDMA
systems, are orthogonal to each other, but the spreaded data related to all the users
may loose its orthogonality, in the channel, in case of high user count. Therefore, the
requirement for alternate mechanism for user separation is needed urgently.
In interleaver-division multiple-access (IDMA) scheme, most of above stated
problems do not exist due to application of user-specific interleavers as alternate way
of user separation in place of unitary spreading PN-sequences used in CDMA scheme.
With IDMA scheme, user separation is achieved with the help of user-specific
interleavers, having low cross-correlation amongst them [76]. As the spreaded user
data is fed to the user-specific interleavers, it results in better orthogonality between
9

resultant interleaved data in the channel. The condition of orthogonality is maintained
for removing the risk of collision between the interleavers [91] in the channel.
In IDMA scheme, orthogonal interleavers are referred as the heart of the
systems. If random interleavers are employed as the means of user separation
[76] in IDMA systems, it results in heavy memory requirement for storing the
user-specific interleavers at transmitter and receiver ends. In [104], power
interleavers were introduced which solves the problem of memory requirement
but increases computational complexity during estimation and turbo processing
[51, 62] of interleavers and deinterleavers, at the receiver end. This interleaver is
named as master random interleaver (MRI). Many more interleavers are reported
in literature but most of them are based on methodology of selection of user-
specific interleavers amongst available random interleavers. Still, the problem of
computational complexity is unresolved for the interleaving and de-interleaving
mechanism in IDMA scheme.
In this thesis, an optimum interleaver is proposed for not only solution of
the problem related to computational complexity and bandwidth requirement but
also for optimization for memory requirement at transmitter and receiver ends.

1.4. Problem Statement

Keeping in mind the problems related to implementation of random and master
random interleavers, it becomes mandatory to look for alternate interleaving
mechanism for the solution of the stated problems. In present work, an attempt has
been made to study various aspects of optimum interleaving mechanism for
interleave-division multiple-access scheme, a potential candidate for next-generation
wireless communication. The analysis has been carried out in terms of BER
10

performance, memory requirement, bandwidth requirement, and computational
complexity along with hardware requirement needed for its field-programmable gate-
array (FPGA) implementation. Specifically, the problem undertaken in this thesis can
be stated as follows:
(i) To suggest a user-specific interleaver generation mechanism for IDMA
scheme and to evaluate its performance using analytical modeling and
simulation.
(ii) To examine the performance of the proposed interleaving scheme
under various conditions including its performance comparison with
other interleaving schemes for various parameters.
(iii) To evaluate the performance of IDMA scheme with proposed
interleaving scheme employing maximal ratio combining (MRC)
diversity technique. Also to compare its performance to that of other
interleaving schemes under similar conditions.
(iv) To investigate into the correlation analysis of proposed and other
interleaving schemes to justify the increment in multiple access
interference (MAI) with relative increment in user count.
(v) To study the performance analysis of proposed interleaving scheme
with its implementation on field-programmable gate-array (FPGA)
system for observing the comparative hardware requirements and
timing constraints with that of random and master random interleaving
schemes.



11

1.5. Research Contributions
In this thesis, the emphasis has been given on analysis and design of optimum
interleaver for IDMA scheme which may fulfill all the requirement of
orthogonality as well must be easy to implement.
(a) The mechanism of interleaving with necessary conditions is presented.
Later, block diagram of IDMA scheme is explained including
transmitter and receiver section.
(b) Further, a tree based interleaver (TBI) mechanism [147], for generation of
user-specific interleavers is proposed for interleave-division multiple-
access (IDMA) scheme. The BER performance comparison of the TBI
with random interleaver (RI) and master random interleaver (MRI)
also known as power interleaver [104] is also presented. Apart from the
performance analysis, all the interleavers have also been analyzed
based on memory requirement and computational complexity. Later,
performance evaluation of IDMA scheme with TBI in uncoded and rate
convolutionally coded environment, is presented.
(c) The performance of IDMA is also demonstrated with maximal ratio
combining (MRC) diversity along with random and master random
interleavers with single transmitter two receiver and other
architectures.
(d) The impact of increment in data length on correlation of user-specific
interleavers has also been investigated. It has been revealed that with
increment in user count, the cross-correlation between user-specific
interleavers goes for hike.
(e) Finally, the FPGA implementation of all the interleavers is
12

demonstrated for observing the timing constraints and hardware
requirement. From the analysis, it becomes apparent that tree based
interleaver is an optimum interleaver.

1.6. Thesis Organization
The remaining chapters of the thesis are organized as detailed ahead. Chapter 2
provides preliminaries on interleavers their mechanism and characteristics along with
conditions for orthogonality. In addition to it, the mechanism of interleave-division
multiple-access (IDMA) scheme is also explained for single path and multipath
environments along with literature review.
In chapter 3, proposed user-specific interleaver generation mechanism named
as Tree Based Interleaver (TBI) is presented for IDMA systems [147]. This
interleaver generation mechanism, optimally, removes the problems of extra
bandwidth consumption, excessive memory requirement, and high computational
complexity inherited with other interleaver generation mechanisms. In this chapter,
the performance comparison of IDMA scheme is presented with random interleavers
and master random interleavers in terms of BER performance, memory requirement,
and computational complexity. Also, the performance of IDMA scheme with unequal
power allocation algorithm has been demonstrated along with all the above stated
interleavers.
The chapter 4 demonstrates the analysis of IDMA scheme with maximal ratio
combining (MRC) diversity at transmitter and receiver end along with various
transmitter-receivers architectures. During the performance analysis of MRC diversity
technique, the enhancement in the BER performance of IDMA scheme has been
13

observed and the performance of IDMA systems with tree based interleaver is found
to be similar to that with random interleavers.
In chapter 5, the correlation analysis of all the three interleavers i.e. random
interleaver, master random interleaver, and proposed tree based interleaver has been
performed to analyze the decrement in BER performance with increment in user
count. Further, all the interleavers have been duly implemented on field
programmable gate array (FPGA) for studying their hardware and timing
requirements.
In chapter 6, the conclusion the thesis is presented along with the possible
directions for further research related to interleavers in IDMA systems.






























14

CHAPTER 2
Overview of Interleave-Division Multiple Access
Scheme


2.1. Introduction
Historically, interleaving was employed in ordering block storage on disk-based
storage devices including floppy disk and the hard disk systems. The primary purpose
of interleaving was to adjust the timing differences between the adjacent bits during
transfer of data between computer and storage media. Interleaving was very common
prior to the 1990s, but, later, faded due to availability of high speed processors. Now-
a-days, all the modern disk storage systems are not at all being interleaved.
In communication systems, interleaving is referred to be technique commonly
used to overcome correlated channel noise such as burst error or fading [8, 9, 15]. In
interleaving mechanism, the input data rearranges itself such that consecutive data bits
are split among different blocks and is swapped in a known pattern amongst them. At
the receiver end, the interleaved data is arranged back into the original sequence with
the help of de-interleaver. As a result of interleaving, correlated noise introduced in
the transmission channel appears to be statistically independent at the receiver and
thus allows better error correction.

2.2. Interleavers in Digital Communication
Interleaving has been frequently used in digital communication and storage systems to
improve the performance of forward error correcting codes. Many communication
channels which are not memoryless in nature, errors typically occur in bursts rather
15

than independently. If the number of errors within a code word exceeds the error-
correcting code's capability, it fails to recover the original code word. Interleaving
ameliorates this problem by shuffling source symbols across several code words,
thereby creating a more uniform distribution of errors.
Typically, in the analysis of modern iterated codes, including turbo and low-
density parity-check (LDPC) codes, independent distribution of errors is assumed
[140]. Systems using LDPC codes therefore typically employ additional interleaving
across the symbols within a code word.

For turbo codes, an interleaver is an integral
component of the architecture, and its proper design is crucial for its good BER
performance. However, application of interleaving increases latency accordingly. This
happens due to requirement of entire interleaved block at the receiver side to recover
the critical data. In multi-carrier communication systems, additional interleaving
across carriers may be employed to mitigate the effects of prohibitive noise on a
single or few specific carriers.
In next section, the mechanism of interleaving process will be duly studied
along with condition of orthogonality.

2.3. Interleavers in IDMA Scheme
The user-specific interleavers play vital role in the efficiency of IDMA system. It not
only provides decorrelation between adjacent bit sequences as provided in the case of
orthodox turbo coding and decoding, but also facilitates a means for decorrelating
various users [91]. The correlation between interleavers should measure how strongly
signals from other users affect the decoding process of a specific user [102]. The
better the interleaver decorrelation, the lesser the iterations, required for detection in
multiuser detection (MUD) mechanism [105]. The decorrelation among the user-
16

specific interleavers provides a mean to reduce the multiple access interference (MAI)
from other users thus helping in the convergence of detection process.
A set of interleavers is considered to be practical if it is easy to generate, and
no two interleavers in the set collide. The transmitter and receiver need not store or
communicate many bits in order to agree upon an interleaving sequence. It may be
demonstrated that a properly defined correlation between interleavers can be used to
formulate a collision criterion, where zero cross-correlation (i.e., orthogonality)
implies no collision [91].
In case of IDMA systems, the transmitter is required to transmit the interleaver
matrix consisting of interleaving pattern along with spreaded data related to users, to
the receiver. So, greater the size of the interleaver, the more bandwidth and resources
are consumed during transmission. Also, it is worth to be mentioned that greater the
size of interleaver, more the orthogonality is achieved amongst interleaver [104].
For better understanding of interleaving mechanism, in the next section, interleaving
process will be discussed.

2.4. Mechanism of Interleaving Process
The interleaver is termed as a mechanism which rearranges the ordering of a data
sequence by means of a deterministic bijective mapping. Let
0 1 1
[ , ,...... ]
N
C c c c

= be a
sequence of length N [109]. An interleaver maps C onto a
sequence such that X is a permutation of the elements of C.
Considering C and X as a pair of N-dimensional vectors, there is one-to-one
correspondence between each element of C and each element of X, as shown
in Figure 2.1.
0 1 1
[ , ,...... ]
N
X x x x

=
i j
c x
17

0 1 2
c

c c
3
c
4
c
5 6
c c




0
x

1
x

2
x
3
x
4
x 5
x
6
x
Figure 2.1: Mechanism of Data Interleaving

An interleaver can then be defined by the one-to-one index mapping function.
( ) : [ A A j i], t t = where A={0,1N-1} for , i j A, e .(2.3.1)
The term i' and j are indices of an element of the original sequence c and the
interleaved sequence x, respectively. The mapping function can be expressed as an
ordered set called interleaving vector t :
[ [0], [1],...., [ 1]]. N t t t t =
k
. (2.3.2)
The element of the permuted sequence X is
th
k
[ ] k
X C
t
= ..... (2.3.3)
The inverse interleaver, i.e. the deinterleaver, restores the permuted sequence
to its original order. Throughout the thesis, the terms t and
1
t

are used to
denote the interleaving and deinterleaving vectors, respectively. With the proper
deinterleaver, the permuted elements can be shifted back to their original positions:
1 1
[ [ ]] [ [ ]] k k t t t t

= k =
n
..(2.3.4)
Replacing k by in (2.3.3), then from (2.3.4),
1
[ ] n t

1 1
[ ] [ [ ]]
.
n n
X C
t t t

= = C . (2.3.5)
18

Since the separation of users is achieved by user-specific interleavers, an
obvious interleaver design criterion is that every two interleavers related to specific
users out of a set of interleavers collide up to minimum extent. With increment in
cross-correlation amongst the interleavers, number of collisions also increases,
resulting in increment in bit error rate (BER) of the system.
The property of minimum collision amongst user-specific interleavers depends
on property of orthogonality. It is referred as an important factor in generating the
interleavers. If the orthogonality is not maintained amongst the user-specific
interleavers, the correlation between the users increases proportionally, with
increment in user count resulting in lower BER performance. Therefore, interleavers
generated according to the orthogonality criteria having minimum number of
collisions are accepted as part of IDMA systems.
Two interleavers
i
t and
j
t (where
i
t

j
t ) are said to be orthogonal, if for
any two words, a and b, the correlation C is such that [91], (.)
( , , , ) ( ( )). ( ( )) 0
i j i j
C a b f a f b t t t t = = .(2.3.6)
Above condition is for the orthogonality check of the user-specific interleavers.

2.5. Interleave Division Multiple Access (IDMA) Scheme
In [76], the interleaver based multiple-access scheme has been studied for high
spectral efficiency, improved performance and low receiver complexity. This scheme
relies on interleaving as the only means to distinguish the signals from different users
and hence it has been called interleave-division multiple-access (IDMA). As reported
in [76, 77], IDMA systems inherits many advantages from CDMA systems, in
particular diversity against fading and mitigation of the worst-case other cell user
interference problem. Furthermore, it allows a very simple chip by chip iterative
19

multiuser detection (MUD) strategy. The normalized MUD cost (per user) is
independent of the number of users [76].

2.5.1. Comparison of CDMA & IDMA Schemes
Comparing CDMA and IDMA mechanisms [129, 67, 68], it is observed that IDMA
scheme may be considered as a special case of Code-Division Multiple Access
(CDMA) scheme. The differences and commonalities of conventional CDMA with
channel coding (top part) and IDMA (bottom part) is presented in figure 2.2.


Figure 2.2: CDMA scheme vs. IDMA scheme

Visualizing the apparent difference between CDMA and IDMA schemes, it is
observed that ordering of interleaving and spreading is reversed in IDMA scheme. In
conventional CDMA scheme, the spreader is user specific, whereas in IDMA scheme,
the interleaver is user-specific. In addition to it, in IDMA scheme, forward error
correction (FEC) and spreading is combined in a single encoder (ENC), which is same
or different for all users [76]. As a consequence, very low rate encoding is used for
forward error correction coding. The spreader has no fundamental relevance any more
in IDMA systems. The spreading of data may be used to simplify the overall encoder
[71]. The interleaver is interpreted as a keyword and only authorized receivers are
FEC
FEC
Spreader
t
m
t
m
Spreader
m
x
m
x
CDMA Scheme
Data Input
IDMA Scheme

Data Input
ENCODER AND SPREADER SECTION
20

able to decode the message [76]. However in [78], scheme exploiting the advantages
of CDMA and IDMA schemes has also been proposed. It is suggested in the approach
that for lower count of users, CDMA may be used while for higher user count, IDMA
scheme may be utilized.
In the next section, the block diagram of IDMA communication system
incorporating transmitter and receiver section is explained.

2.5.2. IDMA Transmitter & Receiver
The upper part of Figure 2.3 demonstrates the transmitter structure of the IDMA
scheme under consideration with K simultaneous users. The input data sequence
of user-k is encoded based on a low-rate code C, generating a coded sequence,
k
d
| |
(1), (2),....... ( )........ ( )
T
k k k k k
C c c c j c J = (2.4.1)
where J is the frame length.
The elements in are referred to as coded bits. Then is permutated by an
interleaver
k
c
k
c
k
t , hence, producing | |
(1), (2),....... ( ) ( )
T
k k k k k
X x x x x J =
k
........ j . Following
the CDMA convention, the element in x will be denoted as chips. Users are solely
distinguished by their interleavers, hence the name interleave division multiple access
(IDMA) scheme [71].
The key principle of IDMA is that the interleavers{ }
k
t , opted for user
separation, should be orthogonal for all the users. It is assumed that the interleavers
are generated independently and randomly. The randomly generated interleavers
disperse the coded sequences so that the adjacent chips are approximately
uncorrelated, facilitating the simple chip-by-chip detection scheme as discussed
below.
21



Figure 2.3: IDMA Transmitter and Receiver Structure

Adopting an iterative sub-optimal receiver structure, as demonstrated in figure
2.3, consisting of the primary signal estimator (PSE) and K single user a posteriori
probability (APP) decoders (DECs), the data is iterated for pre-decided number
iterations before finally taking hard decision on it.

Figure 2.4: IDMA transmission in single path

For single path propagation, as shown in figure 2.4, there is only one path for
the transmission. The multiple access and coding constraints are considered separately
22

in the PSE and DECs. The outputs of the PSE and DECs are extrinsic log-likelihood
ratios (LLRs) about ( ) { }
k
x j , assuming y as the output of PSE and DEC with x as
input for analysis purpose, is defined below as,;
( )
( | ( )) 1
( ) log , ,
( | ( )) 1
k
k
k
p y x j
e x j K j
p y x j
| | = +
=
|
=
\ .

)
.. (2.4.2)
Those LLRs are further distinguished by subscripts, i.e.,
and , depending on whether they are generated by the PSE or DECs.
( ) ( )
PSE k
e x j
( ( )
DEC k
e x j
For the PSE section, y in (2.4.2) denotes the received channel output while for
the DECs, y in (2.4.2) is formed by the deinterleaved version of the outputs of the
primary signal estimator (PSE) block. A global turbo type iterative process is then
applied to process the LLRs generated by the PSE and DECs blocks [76].

2.5.2.1. Basic Primary Signal Estimator (PSE)
Assuming that the channel with no memory and after chip matched filtering, the
received signal from K users, for single path as shown in figure 2.4, can be written as;
j = 1, 2 J ..... (2.4.3)
1
( ) ( ) ( ),
K
k k
k
r j h x j n j
=
= +

where is the channel coefficient for user-k and {n (j)} are samples of an AWGN
process with zero mean and variance, . Assuming that the channel
coefficient { } are known a priori at the receiver. Due to the use of random
interleaver {
k
h
2
0
/ 2 N o =
k
h
k
t }, the PSE operation can be carried out in a chip-by-chip manner, with
only one sample used at a time. ( ) r j
( ) ( ) ( )
k k k
r j h x j j = + .. .. (2.4.4a)
where,
23


' '
'
( ) ( ) ( ) ( ) ( )
k k k k k
k k
j r j h x j h x j n j
=
= = +

... (2.4.4b)
( )
k
j is the distortion (including interference-plus-noise) in ( ) r j with respect to
user-k. From the central limit theorem, ( )
k
j can be approximated as a Gaussian
variable, and ( ) r j can be characterized by a conditional Gaussian probability density
function;
( )
2
( ) ( ( ( )))
1
( ( ) / ( ) 1) exp
2 ( ( )) 2 ( ( ))
k k
k
k k
r j h E j
p r j x j
Var j Var j

t
| |
+
= = |
|
\ .
..(2.4.5)
where E (.) and Var (.) are the mean and variance functions, respectively.
The following is a list of the PSE detection algorithm based on (2.4.4a) ~
(2.4.5), assuming that the a priori statistics ( ) ( ) { }
k
E x j and ( ) ( { )}
k
Var x j are
available [76]. Based on [76], the algorithm for chip-by-chip detection will now be
presented in next sub-section.

2.5.2.2. Algorithm for chip by chip Detection
Step (i): Estimation of Interference Mean and Variance
( ) ( ) ( )
( ) ,
k k
k
E r j h E x j =

. .. (2.4.6a)
( ) ( ) ( ) ( )
2
2
,
k k
k
Var r j h Var x j o =

+ ..... (2.4.6b)
( ) ( ) ( ) ( ) ( ) ( )
,
k k
E j E r j h E x j =
k
.... (2.4.6c)
( ) ( ) ( ) ( ) ( ) ( )
2
.
k k
Var j Var r j h Var x j =
k
... (2.4.6d)
Step (ii): LLR Generation
( ) ( )
( ) ( ) ( )
( ) ( )
2 . .
k
PSE k k
k
r j E j
e x j h
Var j

=
. ... (2.4.6f)
24

2.5.2.3. DEC Function
The DECs in figure 2.3 carry out APP decoding using the output of the PSE as the
input. With binary phase shift keying (BPSK) signaling, their output is the extrinsic
log-likelihood ratios (LLRs) { } ( ( ))
DEC k
e x j of ( )
k
x j defined in (2.4.2), which are used
to generate the following statistics,

,....(2.4.7a)
.. (2.4.7b) ( ) ( ) tanh( ( ( )) / 2),
k DEC k
E x j e x j =
. .....(2.4.7c)
2
( ( )) 1 ( ( ( )))
k
Var x j E x j =
k
In the iterative process, PSE and DEC-k exchange the extrinsic information
about ( )
k
x j . The CBC detection for IDMA scheme can be concluded as follows;
(1) Primary signal estimator generates ( ) ( )
PSE k
e x j by (2.4.6f) for decoder DEC-k.
(2) DEC-k generates ( ( ( )))
DEC k
e x j t , which are used to update mean and variance
of ( )
k
x j .
Under the assumption that { ( )
k
x j } are independent, (2.4.6a)-(2.4.6d) are a
straightforward consequence of (2.4.4a) and (2.4.4b). The Step (ii), shown in
algorithm, is obtained by evaluating (2.4.2) based on (2.4.5). Algorithm shown is an
extremely simplified form for all spreading sequences to be length-1. The operations
in (2.4.6a) and (2.4.6b), i.e., generating E(r(j))and Var(r(j)), are shared by all users,
costing only three multiplications and two additions per coded bit per user. Overall,
the PSE operations shown in step (i) and step (ii), cost only seven multiplications and
five additions per coded bit per user, which is very modest [76]. Interestingly, the cost
per information bit per user is independent of the number of users K. This is
considerably lower than that of other alternatives.
1
( ( ( ))) ( ( ( )))
S
DEC k ESE k
j
e x j e x j t
=
=

t
25


Figure 2.5: Flowchart of decoding mechanism in the receiver of IDMA scheme

The flowchart used for detection mechanism is presented in figure 2.5. The
iterative procedure adopted in the receiver is based on turbo processing [62]. The data
is iterated in the receiver section for the pre-decided number of iterations. After final
iteration in the receiver, the data is decoded with respective mechanism.

2.5.3. IDMA over Multipath Channels
In this section, the basic mechanism of IDMA scheme over multipath channels is
explained. The focus has been kept on the detection algorithm for IDMA scheme in
complex multipath fading channels described in [76]. IDMA scheme, as already
discussed in sub-section 2.5.2, is a recently proposed multiple access scheme, in
which user-specific interleavers are adopted as the only mechanism for user
separation. It can be honored as a particular case of chip interleaved CDMA scheme
[103]. The figure 2.6 shows the multipath transmission for IDMA scheme.
26


Figure 2.6: IDMA in multipath transmission

As so, IDMA scheme inherits many advantages of CDMA scheme including
easy treatment for multipath fading. With the help of random interleaving and chip by
chip (CBC) iterative multiuser detection algorithm, the IDMA scheme is applicable to
cancel MAI and ISI effectively and support systems with large numbers of users in
multipath fading channels.
The upper part of Fig. 2.3 presents the transmitter structure of the IDMA
scheme with K simultaneous users. Let be the data stream of user-k. This data
stream is encoded by a forward error correction (FEC) code, generating a chip
sequence . (Here, the spreaded data is denoted by chip instead of bit as the FEC
encoding may include spreading or repetition coding.) Then is permutated by a
user-specific interleaver-k to produce chip sequence . After quadrature phase shift
keying (QPSK) symbol mapping, the symbol sequence
k
d
k
c
k
c
k
v
k
X =
T
[ ..........; ( );..........; ( )]
k k k
(1); x x j x J is produced, where J is the frame length.
Either superscripts Re and Im or function notations Re(.) and Im(.) are used to
indicate real and imaginary parts, respectively. After symbol mapping, it is observed
that,
Re Im
( ) ( ) . ( );
k k k
h l h j i h j = + (2.4.8)
27

where
Re
( )
k
x j and
Im
( )
k
x j are two coded bits from . The mapping rule for QPSK
used in the simulations is given as follows:
k
v
(0, 0) (1,1)
(0,1) (1, 1)
(1,1) ( 1, 1)
(1, 0) ( 1,1)

(2.4.9)
A multipath propagation channel modeled by a tapped delay-line with several
non-zeros taps is used in the simulation. The fading factor of each tap is Rayleigh
distributed random variable. The power profile and time delay profile can be set. The
channel coefficient can be expressed for several scattered impulses received from L
different paths as . [ (0), (1),.........., ( 1)]
k k k k
h h h h L =
For
1
( )
0
l
l
if
otherwise
t t
o t t
=
=

(2.4.10)
And is the impulse response of path-l with the amplitude
l
h
l
h the phase
l
, the
propagation time delay
l
t and the Doppler frequency shift
, D l
f . As the Doppler Effect
is not taken into account in the simulation, the channel model is simplified for channel
impulse response composed of several scattered impulses received from L different
paths as,
1
( )
0
( ) ( ),
l
L
l
l
h h e

l
t o t t

=
=

....(2.4.11)
As it is assumed that the fading coefficients do not change within one
simulation frame, this channel model is referred as quasi-static Rayleigh fading
multipath channel.
Assuming an L-path channel model with fading coefficients
for user-k, . The received
signal for multipath channel as shown in figure 2.6, can be represented by,
[ (0), (1),.........., ( 1)]
k k k k
h h h h L =
Re Im
( ) ( ) . ( )
k k k
h l h j i h j = +
28

.
( ) ( ). ( ) ( )
k k
k l
r j h l x j l n j =

+ ,(2.4.12)
where are the samples of a complex AWGN process with variance ( ) n j
2
o in each
dimension.
As illustrated in lower part of figure 2.3, an iterative sub-optimal receiver
structure is adopted, which consists of a primary signal estimator (PSE) and K single-
user APP decoders (DECs). The multiple access and coding constraints are considered
separately in the PSE and DECs. In the iterative detection process, the PSE and DECs
exchange extrinsic information in a turbo-type manner [76].
The outputs of the ESE and DECs are LLRs about
Re Im
{ ( ), ( )
k k
} x j x j k (j) as
defined below:
Re
Re
Re
( ( ) 1)
( ( )) log
( ( ) 1)
k
k
k
p y x j
e x j
p y x j
| |
= +

=
\ .
|
|
)
)
,..(2.4.13)
These LLRs are further distinguished by subscripts, i.e.,
and , depending on whether they are generated by the PSE or DECs. For
the PSE, y in (2.4.13) denotes the received channel output. For the DECs, y in (2.4.13)
is the deinterleaved version of the outputs of the PSE. A global turbo-type iterative
process is applied to exchange the LLRs generated by the PSE and DECs, as detailed
below. Here, the focus has been maintained on detecting
Re
( ( )
PSE k
e x j
Re
( ( )
DEC k
e x j
Re
( )
k
x j after receiving r.
(
Im
( )
k
x j can be handled in a similar same way.) The CBC detection algorithm includes
PSE and DEC part as discussed below.
Now, concentrating on the detection of
Re
( )
k
x j for user-k from path-l and
rewrite (2.4.12) as,

,
( ) ( ). ( ) (
k k
r j l h l x j j)
l
, + = + ;.. (2.4.14)
29

where
,
( )
k l
j , consists of the MAI from other users, ISI of the multipath propagation
and the noise. Denote the conjugate of ( )
k
h l by . The received signal is
*
k
h
2
*
,
( ) ( ). ( ) ( ) . ( ) ( ),
k k k
r j l h l r j l h l x j j , + = + = +

k l
. (2.4.15)
where
*
, ,
( ) ( ). ( ),
k l k k l
j h l j , , =

By the central limit theorem,
,
( )
k l
j ,

can be approximated as a Gaussian

variable. This approximation is used by PSE to generate LLR for ( )
k
x j
Im
( r j +
.The phase
shift due to be cancelled out in (2.4.15), which means that is not a
function of
( )
k
h l
Re
(
k
) l
) x j ). Denoting
Re
r j j l ( ) ( l r ) + = + , for simplicity, from 2.4.13,
2
( ) (
( ( ))
k
k
E
j

( ))
,
j
( ( ) ( ) | .
PSE k
r j l
e x j l
Var

) 2 |
k
h = ...(2.4.16)
And
1
0
( ( )) ( ( ))
L
PSE k PSE k l
l
e x j e x j

=
=

..(2.4.17)
where it is assumed that the distortion components in the received samples from
different paths are uncorrelated, so that the LLR values based on individual chips can
be directly summed.
The basic mechanism of rest of the receiver section is similar to that of single
path transmission as explained in sub-section 2.5.2. However, the algorithm is given
below.
Denoting received signal as
, ,
( ) ( ) ( )
k l k k l
r j l h x j j + = + for , 1,... 1 j J L = + +
Then,
'
,
,
( ( )) ( ( ))
k l
k
k l
E r j h E x j l =

.(2.4.18)
2
2
,
,
( ( )) ( ( ))
k l k
k l
Var r j h Var x j l o =

+ (2.4.19)
30

Also,
,
( ( )) ( ( )) ( ( ))
k l k l k
E j E r j l h E x j
,
= + ..(2.4.20)
2
, ,
( ( )) ( ( )) ( ( ))
k l k l k
Var j Var r j l h Var x j = + .(2.4.21)
The output of PSE section is obtained as,
,
,
,
( ) ( ( )
( ( )) 2
( ( ))
k l
PSE k l k l
k l
r j l E j
e x j h
Var j
)

+
= ....(2.4.22)
And
1
0
( ( )) ( ( ))
L
PSE k PSE k l
l
e x j e x j

=
=

,...(2.4.23)
The decoder part performs the APP decoding action using the output of PSE
section as its input.
( ) ( ) tanh( ( ( )) / 2),
k DEC k
E x j e x j =
( ( )) 1 ( ( ( )) / 2)
k k
Var x j E x j =
Since, during the detection process, ESE combines the LLR values related to
all the paths, in RAKE manner, the detection algorithm is referred to as IDMA with
LLR combining (LLRC) in [76].
In the next section, the literature survey related to IDMA systems is presented.
Though, during the survey, overall view of various mechanisms related to researches
and applications have been duly presented, however, more attention have been paid
towards user-specific interleavers, referred as heart of IDMA systems.

2.6. Literature Survey
A conventional CDMA system includes separate coding and spreading operations
during the transmission mechanism of data. Theoretical analysis demonstrates the
optimal multiple access channel (MAC) capacity subject to reservation of entire
bandwidth devoted for coding, suggesting the combination of coding and spreading
31

operations using low-rate codes for maximizing the coding gain. But separation of
users without spreading operation is not feasible in CDMA or in its variant such as
multi carrier CDMA (MC-CDMA) systems [11, 76].
In CDMA scheme, user-specific PN sequences are employed as the only
means of user separation by performing spreading operation on user-specific data [27,
17]. As already stated, CDMA system suffers with lot many problems including MAI,
unsuitability for high-speed burst-traffic, and poor orthogonality of PN codes. So,
keeping in mind the future requirement of wireless communication [5], it becomes
mandatory to look for alternate solution for user separation.
Now, the question arises regarding alternate strategy for user separation. The
possible solutions include narrow band coded-modulation scheme using trellis code
structures [61] and employment of chip-level interleavers [98, 99, 61, 62].
By assigning different interleavers to various users, the improvement in
CDMA scheme has already been reported in [62, 71]. Therefore, the possible solution
to the problem of user separation could be to employ chip-level interleavers [97]. This
principle has been considered previously and its potential advantages have been well
demonstrated [103] showing the possibility of employing interleaving for user
separation in coded systems. For wideband systems, the performance improvement by
assigning user-specific interleavers to various users in conventional CDMA has been
demonstrated in [98]. In [63], study of chip-interleaved CDMA scheme and its
application with maximal-ratio-combining (MRC) technique has been carried out for
combating the problem of inter-symbol interference (ISI) in multiple access channels
(MACs). An interleaver-based user separation scheme, known as interleave-division
multiple-access (IDMA) scheme, has also been reported, in [76, 77] for high spectral
efficiency, improved error performance, and low receiver complexity.
32

The interleave-division multiple-access (IDMA) scheme is a technique that
relies on interleaving as only means for user separation [76]. IDMA not only inherits
lot many advantages from conventional CDMA, including robustness against fading
and mitigation of cross-cell interference, but also accommodates very simple chip-by-
chip (CBC) iterative multiuser detection (MUD) strategy [63] while achieving
impressive bit error rate (BER) performance. The IDMA system with random
interleavers performs similar or even better than its comparable counterpart i.e.
CDMA system. An IDMA system along with randomly and independently generated
interleavers has been presented in [76].
In IDMA systems, there exist several areas which are still open for the
researchers. Few of them includes optimal design of integral component of IDMA
communication system and further application of IDMA mechanism in other areas
including satellite communication, LAN networking, optical communications, power
line communications, MIMO systems, and UWB technologies. In addition to it,
horizons are still open for investigation about optimum modulation, channel coding,
spreading, interleaving, and detection mechanisms.
In [15, 17-19], an overview of CDMA scheme is presented with all the
inherent fruits inside it. In [88], authors have compared orthogonal and non-
orthogonal multiple access schemes and have concluded that multiple access schemes
such as schemes using spread spectrum communication are superior to their counter-
parts.
If multiple transmitters or receivers are employed for data communication, it
results in improvement in performance of system [8]. However, implementation of
multiple-input multiple-output (MIMO) system imposes lot of complexities along
with its optimal performance in [23, 20]. The application of IDMA scheme with
33

MIMO systems has been discussed in [123]. Another evolving technology for IDMA
systems, the cooperative communication technique is also in its developmental stage
and is difficult for its implementation due to its inherent problems including
synchronism amongst users and complexity involved with its transmitters and
receivers. The cooperative relay diversity [138] has also been demonstrated in [125,
126, 124] for IDMA systems.
For orthogonal frequency-division multiple-access (OFDM) systems,
importance of interleaving has been duly identified in [116]. However, its application
with IDMA scheme is explored in [119, 143, 142]. In [119], OFDM-IDMA system
has been analyzed for peak power limitations. In [144], the factor graph based design
of OFDM-IDMA receiver is presented claiming its superior performance in
comparison to conventional receiver. The OFDM-IDMA system shares superior
performance in comparison of OFDM-CDMA systems [143]. Also in [121], the
performance comparison of chip interleaved MC-CDMA with that of MC-IDMA has
been discussed. The simulation results have demonstrated the better BER
performance of MC-IDMA systems to their counterpart for higher user count. As a
candidate of fourth generation mobile communication, the application of IDMA
mechanism has been discussed in [24, 132, 133, 135]. The security issue in IDMA
systems has also been taken up in [129].
The application of space time codes [33, 34, 35] have been duly demonstrated
in IDMA scheme in [71] with its superior performance to other codes. LDPC codes
have also been tested on IDMA scheme in [43, 41, 42], however, the IDMA performs
better with conventional convolutional codes and turbo codes for higher user count
including multipath environment [76, 43]. In [21, 22], chip interleaved based scheme
34

has been demonstrated in ultra wide band (UWB) systems with superior performance.
However, UWB systems are under still developmental stage.
Various detection schemes [57, 51, 137] has also been reported for CDMA
systems in [45, 54, 63, 77]. For IDMA systems, the detection scheme employing turbo
processing based detection algorithm [62] has been demonstrated in [39, 71, 72] for
IDMA systems. In [52], various other low cost multi-user detectors, for IDMA
systems, have been discussed. In [49], blind detection based algorithm has been
presented. In [60], exit chart based detection algorithm has been demonstrated.
Various other detection algorithms have been presented in [58, 59]. In [120, 68],
detection mechanism based comparison between IDMA and CDMA schemes has
been demonstrated. In addition to it, various power allocation algorithms have also
been studied in [80-85, 87-90] for IDMA systems.
Employment of interleavers has also been explored with turbo codes in [95,
96, 29] and convolutional codes in [99]. Various interleaving schemes have also been
studied in the literature for CDMA systems [97-99, 112]. However, in [103],
researchers have reported the interleavers for the purpose of user separation and
demonstrated its superior performance over other similar schemes.
The implementation and performance comparison of various orthogonal and
non-orthogonal interleavers on field-programmable gate-array (FPGA) is also another
area of research. The designing of convolutional interleaver is duly explained in
[146], however, in [145], the FPGA implementation of convolutional interleaver has
been demonstrated. The convolutional interleaver is not termed as orthogonal
interleaver and hence cannot be used in IDMA scheme.
Criteria for a good interleaver design for IDMA include low memory
requirement, easy generation, low correlation among interleavers, and low overhead
35

for synchronization between user and base station [109]. The interleavers used in
IDMA system, are bound to be orthogonal in nature [91]. The orthogonality of
interleavers avoids the risk of collision of interleavers in the system [99]. Various
other conditions for orthogonal user-specific interleavers have been discussed in [91,
105]. Initially, user-specific randomly selected interleavers were employed in IDMA
systems [76]. Various other mechanisms for generation of orthogonal interleavers
have been reported in the literature. However, most of them are based on strategy
related to selection of user-specific interleavers from already calculated interleavers.
If random interleavers are employed for the purpose of user separation, then
lot of memory space will be required at the transmitter and receiver ends for the
purpose of their storage. Also, consideration amount of bandwidth will be consumed
while transmission of all these interleavers. There are also other parameters of interest
with the interleavers including computational complexity at the receiver end. For
solution of problems related to high memory requirement and bandwidth
consumption, researches have taken lot of interest.
Many researchers have proposed the selection of user-specific interleavers
based of shifting mechanism of one randomly selected master interleaver. According
to [92], series of interleavers can be generated by circular shifting a specific pseudo
noise (PN) interleaver, which is generated by a PN sequence generator. In [94, 108],
multiple interleavers are generated by cyclically shifting and self-interleaving a
common mother interleaver in a few steps.
The 2-dimensional interleaver has been proposed in [93] by scrambling the
row indices and column indices of a traditional block interleaving matrix, and
obtaining a master interleaver. Various other user-specific interleavers, in the scheme,
are generated by circular shifting master interleaver. Besides, it is claimed that the
36

minimum distance between two adjacent bits resulted from 2-dimension interleavers
is much larger than random interleaver [93].
Helical interleaver, in [107], has been reported, based on employing helically
shifting pattern on master interleaver for generation of other user-specific interleavers.
Also, in [101], a user-specific interleaver design method based on matrix cyclic
shifting is proposed for lower user count. The interleaver proposed in [100] displays
as advantages of approach as low bandwidth and memory requirements induced from
an algebraic solution and the parallel processing with negligible performance
degradations against random interleavers. It is reported to be specially suitable for
parallel implementation of multiple user detections and decoding of IDMA signals,
resulting in efficient improvements of system throughput.
The progressive edge growth (PEG) algorithm proposed for low-density
parity-check (LDPC) codes is adapted to multi-user interleavers pattern design [106].
It is reported that the designed interleavers perform over random interleavers
especially when the number of users is relative large.
In [104], an interleaver has been proposed which alleviates concerns of extra
bandwidth consumption and memory requirement at transmitter and receiver ends. In
this scheme, user-specific interleavers are generated with the help of master random
interleaver (MRI) and user count at the transmission and receiver ends. During the
transmission, only MRI and the user count need to be transmitted. However, this
interleaver generation mechanism raises the problem of computational complexity
occurring due to computation of user specific interleavers.
In [111], proposed interleaver needs only storage of small number of
parameters related to linear congruences and the transmission of a small number of
bits for synchronization. Users and base stations can derive the interleavers from the
37

identification numbers for different users independently and simultaneously. A
parallel permutation mechanism is also proposed for reducing the generation time of
interleavers.
The interleaver generation mechanism described above solves the memory
requirement problem to store the interleaver patterns at transmitter and receiver ends,
and reduces the amount of information exchange between the mobiles and the base
stations for informing the user-specific interleaving patterns to the receivers.
However, the problem of computational complexity at the receiver is still the matter
of concern.
In this thesis, an interleaver known as tree based interleaver (TBI) is proposed
which optimally alleviates all the matter of concerns for IDMA scheme regarding
user-specific interleavers. A part of this work has already been published in [147].
With this method, not only the interleaver assignment scheme is simplified and
memory cost is greatly reduced, but also the computational complexity required to
generate the interleaving matrix is greatly reduced without sacrificing performance.
Inside this thesis, initially, the mechanism of interleave-division multiple-
access (IDMA) scheme is explained. Further, proposed interleaver named as tree
based interleaver (TBI) is described with its algorithm. A comprehensive study of
performance analysis of IDMA scheme with TBI carried out with comparison to
random interleavers. Later, the interleavers including random interleaver, master
random interleaver, and tree based interleaver, are implemented on field
programmable gate array (FPGA) on Xilinx platform for study of their of hardware
requirement and timing constraints.


38


2.7. Simulation of IDMA scheme

In figure 2.7, the IDMA scheme along with CDMA scheme has been simulated and
compared as reported in literature. During simulation term Eb/N0 has been taken in dB
wherever applicable. In this case, random interleavers have been employed in both of
the cases. In CDMA scheme, randomly selected interleavers are used for conversion
of burst error into random errors, however, in IDMA scheme, randomly selected
interleavers are chosen for user specific identification purpose.
Further, the performance of IDMA scheme is evaluated in AWGN channel
environment. The data length is taken to be 1024 bits with spreading length 16, for 15
iteration, at the receiver side for retrieving the data. The signaling scheme is opted to
be BPSK. The spreading has been performed with repetitive coding common for all
the users. All the simulation has been performed assuming single cell environment
and one transmitter-one receiver architecture. Also, the entire users are assumed to
bear same data lengths.
The simulation results demonstrate the superior performance of IDMA scheme
over its counterpart. Observing for higher user count, the CDMA fails to maintain
orthogonality and shows inferior BER performance in comparison to IDMA scheme
[71]. In figure 2.8, the interleaver taken for simulation is random interleaver (RI)
with Rayleigh channel environment. Initially, for the purpose of simplicity, no
forward error correction (FEC) coding has been added in the scheme. The simulation
results show that with increment in user count, the performance of the scheme
deteriorates. In this simulation, similar power profile has been given to all the users.
Due to application of random interleavers, the memory requirement at the transmitter
and receiver ends is high in order to store all the user-specific interleavers.
39


3 4 5 6 7 8 9 10 11 12
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e


100 user with CDMA
80 user with IDMA
80 user with CDMA
40 user with CDMA
100 user with IDMA
20 user with IDMA
40user with IDMA
20 user with CDMA

Figure 2.7: Simulation of IDMA and CDMA schemes

1 2 3 4 5 6 7 8 9 10
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e
n=1
n=8
n=16
n=24

Figure 2.8: Simulation of IDMA scheme with Random Interleaver

40


2.8. Conclusions
In this section, the concepts of interleavers have been overviewed. Further, a simple
study of IDMA scheme along with its mathematical modeling, have been presented
for single path and multipath conditions. Furthermore, IDMA scheme has been
simulated for AWGN channel with random interleavers, and superior performance of
IDMA scheme has been recorded for similar user count. IDMA scheme has also been
simulated for various user counts and it has been observed that with increment in user
count, the performance of both the schemes is degraded, however IDMA scheme
performs better than its counterpart i.e. CDMA scheme for similar conditions.



















41

CHAPTER 3
Performance Evaluation of Tree Based Interleaver in IDMA
Scheme


3.1. Introduction
Due to their user-specific applications, the interleavers used in IDMA system, are
bound to be orthogonal in nature [91]. The orthogonality of interleavers provides the
means of user separation and avoids the risk of collision of interleavers in the system
[99]. Various other conditions for orthogonal user-specific interleavers have been
discussed in [91, 105]. Initially, user-specific random interleavers were employed in
IDMA systems [76].
If an efficient interleaver has to be designed for IDMA scheme, one has to keep
his attention on lot of criteria including low memory requirement, easy generation,
low correlation among interleavers, and low overhead for synchronization between
user and base station [109]. As reported in section 2.6, various mechanisms for
generation of orthogonal interleavers have been reported [91-94, 106-111]. Most of
the mechanisms, reported in literature are either based on shifting principle of one
randomly chosen master interleaver or choosing user-specific interleavers amongst
available orthogonal interleavers. Interleaver generation mechanism employing
shifting mechanisms are unable to support large user count while the mechanisms
based on selection of user-specific interleavers are bound to avail already generated
buffer of all the random interleavers [32, 111].


42


3.2. Motivation
If random interleavers are employed for the purpose of user separation, then lot of
memory space is required at the transmitter and receiver ends. In addition to it, the
base station (BS) has to use a considerable amount of memory to store user-specific
interleavers, which may cause serious concern, in case of large user count. In [104],
an interleaver known as power interleaver or master random interleaver (MRI), has
been proposed which alleviates concerns of extra bandwidth consumption and
memory requirement at transmitter and receiver ends. In this scheme, user-specific
interleavers are generated with the help of one randomly selected master interleaver
known as master random interleaver (MRI), for available user count at the
transmission and receiver ends. During the transmission, only MRI and the user count
need to be transmitted. Certainly, this interleaver solves the problem of extra
bandwidth consumption and memory requirements at the transmitter and receiver
ends, however, raises an additional problem of computational complexity occurring
due to iterative computation of user specific interleavers. The problem of
computational complexity has not taken care, in most of the interleaving mechanisms
reported in literature. Similar approach has also been reported in [111] by utilizing
small number of parameters related to linear congruences for generation of user-
specific interleavers. However, the problem of computational complexity at the
receiver is still the matter of concern.
The computational complexity also increases, at the receiver end, due to
application of turbo processor [62], which utilizes iterative interleaving and
deinterleaving operations during the process of iterative decoding. Therefore, large
43

amount of calculations are required in receiver section and hence, the computational
complexity is increased drastically in case of large user count.
In this chapter, the proposed Tree Based Interleaver (TBI) is evaluated to alleviate
concerns including bandwidth requirement, memory requirement and computational
complexity optimally. Its performance is also compared with that of random
interleaver (RI) and master random interleaver (MRI) in due simulations performed
under similar environment. The performance of TBI is also evaluated using unequal
power allocation algorithm without sacrificing the performance. All the simulations
have been performed in MATLAB environment.

3.3. Mechanism of Tree Based Interleaver
The Tree Based Interleaver (TBI) is basically aimed to optimize the problems of the
computational complexity and memory requirement which occurs in MRI [104] and
RI [76] respectively. In addition to it, TBI also targets on consumption of optimal
bandwidth during communication process.
In case of TBI generation mechanism, two randomly generated master
interleavers
1
and
2,
are taken initially. These interleavers are bound to have
orthogonality between each other. The zero cross correlation between two randomly
selected interleavers ensures the minimal cross correlation between other generated
user-specific interleavers, using tree based interleaver generation algorithm. The
allocations of the interleaving masks follow the tree format as shown in figure 3.1
demonstrating tree based interleaver mechanism. The interleaving masking diagram is
shown upon fourteen users only for the sake of simplicity. For obtaining the
interleaving sequence of the 14
th
user, the TBI mechanism needs only 2 cycles of
44

clock, as compared to many more cycles needed in case of master random interleaver
method.

14
=
2
(
2
(
2
))


Figure 3.1: Interleaving strategy for Tree Based Interleaving scheme.

As already stated, the mechanism of Tree Based user-specific interleaver
generation is based on two master interleavers
1

and
2
, which are randomly
selected. User specific interleavers are designed using a combination of randomly
selected master interleavers. The interleaver
1
is opted for upper branch while 2


is reserved for initiation for lower branch. Upper branch is selected for the case of odd
user count while lower branch is selected with even user count. First user is allocated
45

with interleaver
1
while the second user, is reserved with the interleaver
2.
In case
of third user, the user-specific interleaver is decided with
1
(
1
) and for fourth user,
the interleaving sequence will be
2
(
1
).
In this manner, large number of users may be allocated with user specific
interleavers with extremely less complexity. Here, for the sake of simplicity, it is
assumed that the entire user bears same data length along with same spread length and
similar channel coding. However, in case of uneven data lengths related to all the
users, the longer data length is selected as master data length and interleavers are
designed based on master data length. If for any user, the data length is less than the
master data length, the specific count of zeros are padded in order to equalize all the
data lengths related to all the users.

Algorithm of Tree Based Interleaver
Step 1: Master interleaver
1
is randomly generated having the length of data block
(data length spreader length).
Step 2: Master interleaver
2
is randomly generated having the length of data block
(data length spreader length).
Step 3: According to user k, level (L) of Tree is determined. Hence, number of users
in that level =K= . 2
L
Step 5: All the possible combination ( 2 ) of interleavers are generated as 1 (
L
1
),
2 (
1
)..and so on.
Step 6: According to the user K, user-specific interleavers based on particular
combination of master interleavers is chosen and thereafter data is interleaved
accordingly.

46

The memory required by the tree based interleaver (TBI) generation method is
only slightly more than that required for master random interleaver (MRI) generation
method [104] due to requirement of two orthogonal interleavers in place of one
interleaver. However, from the TBI algorithm and figure 2.3 related to IDMA
mechanism, it is evident that, bare minimum number of iterations is required for
generation of user-specific interleaving sequence. The TBI generation scheme is
extremely efficient for reduction of computational complexity as compared to that in
MRI generation scheme [104] as shown in figure 3.1. However, the Tree Based
Interleaving scheme reduces the computational complexity that occurs in the MRI
generation scheme.
For the performance evaluation of TBI generation mechanism for the users,
with IDMA systems, similar data length with other parameters including modulation,
channel coding, and decoding mechanism is assumed in MATLAB environment.

3.4. Performance Evaluation of Tree Based Interleaver
In most of simulation presented, in this thesis, the performance of proposed tree based
interleaver is compared with that of random interleaver. The reason for choosing
random interleaver for performance comparison lies in the fact that few user-specific
random interleavers are chosen from the available thousands of interleaving
combinations and the orthogonality between these randomly selected interleavers is
quite assured. In fact, the interleaving combinations are the alternate possible ways of
rearrangement of user data. As an example, if individual user is having data of L bits,
there exist 2L bits after rate convolutional coding mechanisms. As shown in figure
2.3 in section 2.5, the available coded 2L bits are further spreaded with spread length
S. The spreaded data is, now, termed as available 2 chips, which further, LS
47

provides (2 ) { 2 2 2} LS LS combinations for their rearrangement. In fact, the
available (2 ) { 2 2 2} LS LS combinations for rearrangement are the available
counts of interleavers. For further processing, the user-specific data has to be
rearranged in randomly selected user-specific interleavers out of available
combinations. Thus, data arranged in few randomly selected interleavers out of
available counts of interleavers, provide better orthogonality between user-specific
interleaved chips. For only 1024 bits of data length for all the users along with rate
convolutional coding mechanism and spread length 16, the number of possible
interleaving combinations are 53,68,54,528. If few interleaving user-specific
mechanism are randomly chosen out of these combinations, the probability of their
orthogonality with each other is quite high. However, such randomly selected random
interleavers (RIs) require higher bandwidth consumption along with huge memory
consumption at transmitter and receiver ends along with lot of computational
complexity [104].
If the spreaded and interleaved chips, ready for transmission, in case of IDMA
scheme, are compared to that of CDMA scheme, better orthogonality is observed for
IDMA systems for higher user count. The spreaded data of CDMA systems may lose
its orthogonality between them with increment in user count due to poor orthogonality
between unitary sequences used for user separation. However in case of IDMA
systems, the already spreaded data is later interleaved with few randomly selected
interleavers out of available counts of interleavers. This ultimately provides
better orthogonality for interleaved data related to IDMA systems in comparison to
that of for CDMA systems.
2
2
LS
During the transmission from transmitter end, the related interleavers are sent
along with interleaved data. The interleaved data is in form of chips. In case of
48

random interleaver (RI) employed with IDMA scheme, high amount of bandwidth is
required during transmission. Also, there is high memory requirement for storage of
all the user-specific interleavers at transmitter and receiver ends. However, due to
availability of all the user-specific interleavers at receiver end, the problem of
computational complexity does not exist. In case of master random interleaver (MRI)
[104], only master interleaver along with user count is sent along with interleaved
chips, so there is evident reduction in bandwidth consumption and memory
requirement. However, for calculation of user-specific interleavers from MRI, there is
necessity of high computations required at the transmitter and receiver ends. Also,
comparatively lesser memory is still needed to store all the intermediate results [147].
In case of proposed tree based interleaver (TBI) mechanism, two randomly selected
master interleavers are sent along with interleaved data [147]. The tree algorithm at
the receiver end easily recalculates the required user-specific interleaver. Therefore,
the amount of calculation and memory requirement is solved significantly while there
is minor increment in bandwidth consumption in comparison to MRI due to
transmission to two master interleavers in comparison to one master interleaver for
the later.
In this section, the performance of proposed tree based interleaver (TBI) will
be examined in comparison to that of random interleaver (RI) and master random
interleaver (MRI). The assumption made for the simulation includes single cell
environment with one base transmitter to one mobile receiver architecture. The entire
users connected to the base transmitter are assumed to bear same data lengths and
spread lengths including same forward error correction (FEC) coding if used. The
simulations are performed in MATLAB environment.
49

Figure 3.2 presents comparative simulation of TBI, MRI and RI for 2048 bits
of data length for 64 users. The spread length for all the users is taken to be 16 while
the iteration count has opted to be 15, for iterative decoding at the receiver. The
simulation is performed under AWGN environment without any forward error
correction coding with unequal power allocation algorithm mentioned in section 3.5.
If the bit energy Eb, to be transmitted, is in watts and noise power N0 is in watts/ Hz,
then Eb/N0 is decided as a ratio and its unit is in terms of dB. The simulation result
demonstrates the similar BER performance of TBI to that of RI and MRI in case of
similar simulation conditions.
0 5 10 15 20 25
10
-4
10
-3
10
-2
10
-1
10
0
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e


Tree Based Interleaver
Random Interleaver
Master Random Interleaver
Figure 3.2: Performance of Tree based Interleaver with Random Interleaver
The IDMA scheme, inbuilt with random interleaver, imposes the problem of
extra bandwidth consumption in the channel, along with high memory requirement at
the transmitter and receiver ends. Table 3.1 presents the memory requirement of all
the three already discussed interleavers.
In figure 3.3, the simulations results demonstrating comparison of RI, MRI,
and TBI mechanisms for memory requirement, is presented. The simulations were
performed for 100 users with 256 bits of data length along with spread length to be
50

16. The simulation results conclude that the performance of proposed tree based
interleaver is very close to the desired ideal status of the results.

Table 3.1
Comparison of Memory requirement of user-specific interleavers in IDMA scheme
Sr. No. User
Count
Memory
requirement
of Random
Interleaver (RI)
Memory
requirement
of Master
Random
Interleaver (MRI)
Memory
requirement
of Tree Based
Interleaver (TBI)
1. 2 1 1 2
2. 6 6 1 2
3. 14 14 1 2
4. 30 30 1 2
5. 62 62 1 2
6. 126 126 1 2


The result demonstrates that the memory required for storing the user-specific
interleavers is user dependent for random interleavers in case of its utilization in
IDMA scheme, while it is found to be at minimum level, in case of deployment of
master random interleaver [104]. For tree based interleaver, the requirement of
memory is observed to be little bit high in comparison to that required in case of
master random interleaver, however, it is extremely less when compared with
requirement in case of random interleaver.
The stated problems of extra channel bandwidth consumption and high
memory requirement were solved by master random interleaver in [104]. However,
the problem of computational complexity was born with implementation of master
random interleavers in IDMA systems. In table 3.2, the proposed interleaver is
51

examined on the ground of computational complexity with that of master random
interleaver [104] at transmitter end.

Table 3.2
Comparison of Computational complexity of user-specific interleavers/ deinterleavers at
transmitter end

Sr. No. User
Count
Interleaving/
Deinterleaving
complexity with
Random Interleaver
(RI)
Interleaving/
Deinterleaving
complexity with
Master Random
Interleaver (MRI)
Interleaving/
Deinterleaving
complexity with
Tree Based
Interleaver (TBI)
1. 2 1 1 1
2. 6 1 5 2
3. 14 1 13 3
4. 30 1 29 4
5. 62 1 61 5
6. 126 1 125 6


The table 3.2 demonstrates the superior performance of tree based interleaver
over master random interleaver when implemented with IDMA systems. In case of
random interleavers there is almost no computational complexity during calculation of
interleaving/ deinterleaving sequences, but it imposes problems of extra bandwidth
consumption along with problem of high memory requirement at transmitter and
receiver ends. In master random interleaver (MRI), the problem of computational
complexity is too severe at least at transmitter end. The term computational
complexity stands for the count of interleavings/ deinterleavings required for
calculation the user-specific interleaver. The situation becomes critical in turbo
processing mechanism in IDMA receiver in case of large user count. The simulation
of computational complexity related to all three user-specific interleavers is presented
52

in figure 3.4 at transmitter end. The parameters opted for simulation includes 256 bits
of data length along with spread length to be 16 for 100 users.

0 10 20 30 40 50 60 70 80 90 100
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
x 10
6
Number of Users
M
e
m
o
r
y

R
e
q
u
i
r
e
m
e
n
t

o
f

I
n
t
e
r
l
e
a
v
e
r
(
N
o
.
o
f

b
i
t
s

r
e
q
u
i
r
e
d
/
u
s
e
r
)


Memory requirement of Random Interleaver
Memory requirement of Master Random Interleaver
Memory requirement of Tree Based Interleaver
Figure 3.3: Comparison of RI, MRI, and TBI for memory requirement

Simulation results presented in figure 3.4 demonstrates that computational
complexity of master random interleaver increases with increment in user count while
the computational complexity for random interleaver is at minimum level. However,
for tree based interleaver (TBI), the computational complexity is little bit high in
comparison to that needed for random interleaver, but extremely less when compared
to that of master random interleaver. Also, it has to be noted, here, that the BER
performance of all the interleavers is observed to be similar as demonstrated in figure
3.2, while the BER performance of master random interleaver and random interleaver
is already demonstrated to be at similar level in [104].

53

Table 3.3
Comparison of Computational complexity of user-specific interleavers/ deinterleavers at
receiver end

Sr. No. User
Count
Interleaving/
Deinterleaving
complexity with
Random Interleaver
(RI)
Interleaving/
Deinterleaving
complexity with
Master Random
Interleaver (MRI)
Interleaving/
Deinterleaving
complexity with
Tree Based
Interleaver (TBI)
7. 2 1 1 1
8. 6 1 4 2
9. 14 1 6 3
10. 30 1 8 4
11. 62 1 10 5
12. 126 1 12 6

0 10 20 30 40 50 60 70 80 90 100
0
10
20
30
40
50
60
70
80
90
100
Number of Users
C
o
m
p
l
e
x
i
t
y

o
f

i
n
t
e
r
l
e
a
v
e
r

(
N
o
.
o
f

I
n
t
e
r
l
e
a
v
i
n
g
s
/
u
s
e
r
)


With Random Interelaver
With Master Random Interleaver
With Tree Based Interleaver

Figure 3.4: Comparison of RI, MRI, and TBI for computational complexity at transmitter end

54

The computational complexity at the receiver end is found to be at minimum
level due to no need of computations for user-specific interleavers at receiver end.
The computational requirement of all the considered interleavers is presented in table
3.3 at receiver end. The simulation for computational complexity is also demonstrated
in figure 3.5. From the presented table and concerned simulated figure, it is well
demonstrated that, even at receiver end, the tree based interleaver performs better to
that of master random interleaver. However the performance of Tree Based
Interleaver is little bit inferior to that of Random Interleaver.

Figure 3.5: Comparison of RI, MRI, and TBI for computational complexity at receiver end.

If the parameter of bandwidth consumption is compared for RI, MRI, and TBI
mechanisms, it can be easily understood that RI will consume the maximum
bandwidth due to requirement of transmission of the entire user-specific interleavers
while TBI will consume comparatively more bandwidth in comparison to MRI, due to
55

deployment of two interleavers in place of only one interleaver for deployed in case of
MRI. The formats of data required for transmission are presented in figure 3.6.

Figure 3.6: Data formats of RI, MRI, and TBI mechanisms in IDMA scheme

The BER performance of TBI in IDMA scheme is demonstrated in figure 3.7
with AWGN channel with variation in user count. The simulation has been observed
for 128 bits of data length for all the users. The iterations required in receiver section
are chosen to be 15, while spread length is taken as 16.
Figure 3.8 presents the BER performance of IDMA scheme in AWGN channel
and flat Rayleigh fading channel environment with TBI. During simulation, the bit-
energy to noise (Eb/N0), has been taken in dB wherever applicable. The term Eb is the
bi-energy entering into the IDMA system. The data length chosen for simulation is
128 bits with 15 iterations count along with 16 spread length. The result demonstrates
that the performance of IDMA systems in Flat Rayleigh fading environment is
inferior to that of AWGN channel environment for similar simulation conditions. The
BER has been calculated by comparing the status of transmitted & received bit at the
receiver side.
56



0 5 10 15 20 25 30
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
Eb/No(dB)


n=1
n=8
n=16
n=24
n=32
n=48
n=50
n=56
n=64

Figure 3.7: Simulation of TBI in multi-user environment
0 5 10 15 20 25
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
Eb/No(dB)
B
i
t

E
r
r
o
r

R
a
t
e


user 1 Flat Rayleigh fading channel
user 1 AWGN channel
user 16 Flat Rayleigh Fading
user 8 AWGN channel
user 8 Flat Rayleigh Fading
user 16 AWGN channel
user 48 AWGN channel
user 48 Flat Rayleigh Fading channel

Figure 3.8: Uncoded IDMA scheme in AWGN and Flat Rayleigh fading environment
57


Further, the performance of RI is simulated with rate convolutional coding
with AWGN channel environment. The other parameters including spread length, data
length, are opted to be 16, and 128 respectively. However, the user count is chosen as
32 for the purpose of simulation. The equal power allocation mechanism is adopted
for all the users. The simulation results presented in figure 3.9, demonstrates superior
performance of rate convolutional coded IDMA systems when compared to un-
coded IDMA systems. The results confirm that coded communication system
performs always better than its uncoded counterpart [76].

2 4 6 8 10 12 14 16
10
-7
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
E b/No
B
i
t

E
r
r
o
r

R
a
t
e


32 user uncoded RI
32 users coded RI
Figure 3.9: Simulation of RI in coded and uncoded IDMA scheme
However, for the same parameters of simulation, the results for uncoded
IDMA systems in AWGN environment comparing the performance of random
interleavers and tree based interleavers are presented in figure 3.10 for variation in
user count. The simulation results demonstrate almost similar BER performance of
IDMA system with RI to TBI.
58

0 5 10 15 20 25
10
-7
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e


Tree Based Interleaver 16 users
Random Interleaver 16 users
Random Interleaver 32 users
Tree based Interleaver 32 users
Random Interleaver 48 users
Tree Based Interleaver 48 users
Random Interleaver 64 users
Tree Based Interleaver 64 users
16 users
32 users
64 users
48 users

Figure 3.10: IDMA scheme in uncoded environment for variation in user count


2 4 6 8 10 12 14 16
10
-7
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
E b/No
B
i
t

E
r
r
o
r

R
a
t
e


without coding 32 users TBI
with coding 32 users TBI
Figure 3.11: Simulation of RI in coded and uncoded IDMA scheme
59

0 5 10 15 20 25
10
-7
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e


Random Interleaver with 32 users
Tree Based Interleavers with 32 users
Random Interleaver with 32 users
Tree Based Interleaver with 16 users
Tree Based Interleaver with 48 users
Random Interleaver with 48 users
Random Interleaver with 64 users
Tree Based Interleaver with 64 users
48 users
64 users
16 users
32 users

Figure 3.12: IDMA scheme in coded environment for variation in user count


In figure 3.11, TBI has simulated with similar parameters chosen for
simulation of RI in coded and uncoded environment. The simulation result
demonstrates the anticipated results with significant improvement in performance of
rate convolutional coded IDMA scheme. The result demonstrates the decrement in
performance with increment in user counts which ultimately happens due to
increment in multiple access interference. The increment in MAI with hike in user
count has been duly explained in chapter 5.
In figure 3.12, the IDMA scheme has been simulated for various user counts
with rate convolutional coding environments with similar simulation parameters.
Simulation results in figure 3.10 and 3.12, confirm the similar performance of tree
based interleaver to that of random interleaver, however the decrement in BER
performance has been observed for both the interleavers with increment in user count.


60


3.5. TBI with Unequal Power Allocation Algorithm
In wireless communication, direct-sequence code-division multiple-access (DS-
CDMA or simply CDMA) has been suitably adopted for second and third generation
cellular mobile standards. In [85, 88], it has been shown that multi-user detection
together with unequal power control can enhance the performance of CDMA systems.
Unequal power allocation strategy is a technique to further enhance system
performance in multiple access techniques. It has also been demonstrated in [88] that
IDMA together with CBC algorithm can achieve performance close to the theoretical
limits.
In this section, the unequal optimized power allocation technique [80] is used
with random interleavers & tree based interleavers for the purpose of performance
comparison to reach to the conclusions in IDMA system. However, any modest power
allocation algorithm may be applied with IDMA scheme employing TBI for
improvement in BER performance.

3.5.1. UnequalPowerAllocationMechanism
Considering the k-user IDMA system, shown in figure 2.3, presented in section 2.5, it
is assumed that all the user bear similar simulation conditions. At the transmitter for
user-k, the information bit stream is first encoded by an FEC encoder and
then interleaved and transmitted over a multiple access channel (MAC) with proper
power control factor . The j-th chip of the length J signal sequence is denoted by
k
d ( )
k
ENC
k
P
( ) { } 1, 1
k
x j e + , after encoding and interleaving at the transmitter for user-k.
61

For simplicity, it is assumed that, for all users, same forward error codes (FEC) is
used and the same BER performance is required with BPSK signaling in AWGN
channel.
In view of unequal power allocation [80], equation (2.4.3) can be rewritten for
the received signal as;
. (3.4.1) ( ) ( ) ( )
1
, 1, 2, .....
K
k k k
k
r j P h x j n j j J
=
= + =

where is the channel coefficient for user-k, and

k
h ( ) { }
n j are samples of an additive
white Gaussian noise (AWGN) process with zero mean and variance,
[76].
2
/ 2
o
N o =
Focusing on user-k, equation (3.4.1) may be rewritten as,
( ) ( ) ( ) .
k k k k
r j P h x j j = + (3.4.2)
where,
( ) ( ) ( ) ( ) ( ) . .
k k k k k i i
i k
j r j P h x j P h x j n j
=
= +

.. . (3.4.3)
is the noise-plus-interference component in ( ) r j in (17) with respect to ( )
k
x j .
Based on the assumption that ( )
k
j is a Gaussian random variable with mean
and variance ( ) (
k
E j
) ( ) (
k
Var j
)
, output of the PSE for user-k as follows;
( ) ( )
( ) ( ) ( ) ( )
( ) ( )
2 .
k k k
PSE k
k
P h r j E j
e x j
Var j

= (3.4.4)
Let
( )
{ }
l
k
is the average signal to noise and interference ratio (SNIR) for the
outputs of the PSE after the l-th iteration &
( )
( )
l
k
f be the average variance of the
outputs of driven by an input sequence with SNIR
k
DEC
( ) l
k
[80]. Also, be the
power control factor in the scheme.
k
P
62

( ) l
k
can be approximately tracked using the following recursion for AWGN channel;
( )
( )
( )
2
1
2
2
, , 0,1,.... 1
l k k
k
l
i i i i
i k
P h
k l L
P h f

o
+
=
= =
+

................. (3.4.5)
where L is the maximum number of iterations along with as power related to
user.
k
P
th
k
For AWGN channels, it is assumed that
2
1,
k
h k = . Then recursion (3.4.5)
reduces to the following form;
( )
( )
( )
1
2
, , 0,1,....( 1),
l
k
k
l
i i
i k
P
k l L
P f

o
+
=
= =
+

.. (3.4.6)
The algorithm, for optimization of total power
k
k
P

while achieving the required

performance for
( )
,
l
k
k > I , with specifiedI, is presented in [80].

3.5.2. NumericalResults
The optimization target is to minimize the required
b o
E N to achieve BER
4
10

s .
The simulation is performed using random interleaver and tree based interleaver.
Assumptions are made of similar data length and spread length for all the users to
employ same forward error correction coding if used.
In figure 3.13 and 3.14, the results demonstrate the variation in BER
performance with variation in data lengths user for specific IDMA systems for RI and
TBI respectively.
63

2 4 6 8 10 12 14 16
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e


datalength=128
datalength=256
datalength=512
datalength=1024
Figure 3.13: Simulation results for 32 users without coding with RI with various data lengths

2 4 6 8 10 12 14 16
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e


datalength=128
datalength=256
datalength=512
datalength=1024

Figure 3.14: Simulation result for 32 users without coding with TBI with various data lengths

64

In the simulations, iteration count has been fixed on 15 in the receiver section
while data length has been varied from 128 bits to 1024 bits. The spread length is
taken as 16 while no FEC coding has been considered. The results expresses similar
BER performance of tree based interleaver when compared with that of random
interleaver. Also, it is observed that, the BER performance of the system improves
with increment in data lengths for users. It happens due to availability more random
interleavers for increment in data lengths.

0 5 10 15 20 25
10
-8
10
-7
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e


Tree Based Interleaver 16 users
Random Interleaver 16 users
Random Interleaver 32 users
Tree based Interleaver 32 users
Random Interleaver 48 users
Tree Based Interleaver 48 users
Random Interleaver 64 users
Tree Based Interleaver 64 users
32 users
16 users
48 users
64 users
Figure 3.15: IDMA scheme in uncoded environment for variation in user count with RI and
TBI with unequal power allocation algorithm

Figure 3.15, presents the simulation results for IDMA scheme using random
interleavers with 256 bits of data length for all the users along with spread length to
be 16. The iterations at the receiver are chosen to be 15. During the simulation, the
unequal power allocation algorithm [80] has been adopted. The simulation is
performed without any forward error corrections (FEC) coding. The outcome of the
65

simulation depicts similar BER performance of tree based interleaver to that of
random interleaver. For the identical simulation parameters, similar performance has
been observed in figure 3.16 for rate convolutionally coded environments.
From both of the simulations, it is evident that the system performs better in
terms of BER performance for FEC coded condition with unequal power allocation
algorithm. Also, with increment in user count the system performance goes for dip
due to increment in multiple access interference (MAI).
0 5 10 15 20 25
10
-9
10
-8
10
-7
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e


Tree BasedInterleaver 32 users
Random Interteaver 32 users
Tree BasedInterleaver 16 users
Random Interleaver 16 users
RandomInterteaver 48 users
Tree Based Interteaver 48 users
Tree Baed Interteaver 64 users
Random Interteaver 64 users
32 users
48 users
16 users
64 users
Figure 3.16: IDMA scheme in rate convolutionally coded environment for variation in
user count with RI and TBI with unequal power allocation algorithm

3.6. Conclusions
In this chapter, the proposed tree based intreleaver (TBI) is discussed along with
simulation in IDMA environment. The performance of TBI in terms of BER, memory
requirement, and computation complexity, is compared with those for random
interleaver (RI) and master random interleaver (MRI). From the simulations, it is
66

observed that the BER performance for all the interleaver is similar under same
environment while TBI outperforms MRI in terms of computational complexity. TBI
also performs better to RI in case of memory requirement while requires little bit of
higher memory if compared to that in case of MRI. So, it may be concluded that TBI
performs optimally in comparison to all the interleavers.
Later, the performance of TBI is also compared to that of RI under influence of
unequal power allocation algorithm. Here, it is worth mentioning that IDMA scheme
is anticipated to perform better with any modest power allocation algorithm when
implemented with TBI in comparison to other discussed interleavers.
Therefore, it can be concluded that IDMA systems performs optimally when
implemented with TBI mechanism.












67

CHAPTER 4
Performance Evaluation of Tree Based Interleaver in IDMA
Scheme with Maximal Ratio Combining (MRC) Diversity
Mechanism

4. 4.1. Introduction 1. Introduction
Diversity is a powerful technique for communication receivers which improve the
wireless link at a relatively low cost. They are employed to primarily improve
performance over a fading radio channel [8]. In general, fading is classified into small
scale and large scale fading. This notion is also related to coherence time of the
channel which measures the time over which fading process is fully correlated. In
general, the fading is said to be slow if the symbol period is less than the coherence
time of the channel, otherwise it is considered to be fast fading. In slow fading,
particular fade will affect successive symbols while in fast fading the fading is
independent from symbol to symbol.

4.2. Diversity Mechanisms
The effective techniques for combating fading problem are based on various
parameters including time, space, and frequency. Time interleaving, together with
error correction coding, can provide diversity improvement for the signal. However,
it results in large time delays in case of slowly varying channel. Equivalently, spread
spectrum techniques are ineffective when the coherence bandwidth of the channel is
larger than the spreading bandwidth or, equivalently, where there is relatively small
delay spread in the channel [8, 139].
68

In a wireless communication system, if the antenna elements of the receiver
are separated by a fraction of the transmitted wavelength, various copies of the
information signal, can be combined suitably to get the received signal. Such a
diversity technique is termed as Antenna or Space diversity technique. Diversity
techniques are effective when the branches considered are assumed to be
independently faded or the envelopes are totally uncorrelated.
In most scattering environments, antenna diversity is a practical, effective and,
hence, a widely applied technique for reducing the effect of multipath fading. The
classical approach is to employ multiple receiving antennas and perform combining or
selection and switching mechanisms in order to improve the quality of the received
signal. Now-a-days, another diversity method, known as Transmit Diversity, is
becoming popular. In this technique multiple transmitting antennas and single receive
antenna combination is added to improve the reception quality of all the remote units
in coverage area of particular cellular system [113, 139].
In next section, the major techniques related to diversity mechanisms will be
discussed in brief.

4.2.1. Frequency diversity
The diversity may also be achieved by separating the modulated signal with M
different carriers with frequency gap of at least equal to coherence bandwidth of the
channel. With this approach, it is ensured that the M different copies of the signal
undergo fading independently, in the channel.
Further, on the receiver side, the various independently faded received signals
are combined optimally with help of maximal ratio combiner (MRC) to give a
statistics for the decision on received data. The MRC combiner will be discussed
69

separately later in section 4.3.2. Frequency diversity can be used to combat frequency
selective fading [8, 139, 113].

Figure 4.1: Frequency diversity

4.2.2. Time diversity
The diversity can also be achieved by transmitting the desired signal during M
different time slots separated by at least the coherence time of the channel, i.e., each
symbol is transmitted for M times, as shown in figure 4.2.


Figure 4.2: Time Diversity

The time intervals between transmissions of the same symbol are adjusted to
be at least equal to the coherence time so that different copies of the same symbol
70

undergo independent fading in the channel. The transmission of same symbol for
multiple times provides more probability of getting better SNR at the receiver when
combined optimally using suitable combining scheme. For combating time selective
fading, the error control channel coding, can be combined with interleaving [10], [12].

4.2.3. Space diversity Mechanism
In this approach, the diversity is achieved by using multiple receiving antennas for M
copies of the transmitted signal and vice versa. The antennas should be spaced far
enough apart so that multiple copies of the signal go through independent fading in
the channel. In this case, unlike other diversity techniques, no additional work is
required at the transmission end, and no additional bandwidth or transmission time is
required. However, physical constraints including antenna placement, may limit its
applications in the field, practically. This diversity can be employed for combating
both frequency selective fading and time selective fading in the channel.
Space diversity technique is based on orthogonal design and is referred to be
optimum when there is no interference at all. In the presence of interference, both the
useful signal and the interference may be added coherently at the receiver [113]. This
implies that the diversity technique potentially increases the experienced interference
compared to a single antenna systems (SAS).
The most commonly used space diversity method is receiver diversity, where
several uncorrelated replicas of the signals are combined at the receiver in order to
improve the signal reconstruction. Diversity can also be achieved through transmitter
diversity, by using several antennas to transmit the same signal. These antennas are
placed sufficiently far apart to get uncorrelated signals at the receiver end. A simple
two-branch transmit diversity scheme, referred to as space-time multiplexing,
71

provides the same diversity order as MRC with one transmit antenna and two receive
antennas, without increasing the computation complexity [113]. Transmitter diversity
is not commonly used in practice yet.
In general, in communication systems, implemented with suitable diversity
mechanism, the receiver is provided with multiple copies of the similar signal which
are transmitted over multiple communication channels. The basic idea of diversity
mechanism is repetition or redundancy of information. In virtually all the
applications, the diversity decisions are made by the receiver and are unknown to the
transmitter [139].
In next section, the concepts of transmit and receive diversity mechanisms will
be discussed in brief.

4.2.3.1. Transmit diversity
In transmit diversity, as shown in figure 4.3, there are a multiple transmitting antennas
transmitting the same signal to single receiver which receives the various signal
copies obtained through the various independent paths.

Figure 4.3: Transmit diversity with multiple antennas at transmitter side

Here, different signal copies experience different degrees of fading. Some of
the signal copies can be deeply faded whereas some can have sufficient signal energy
72

to be detected. The detection decision is made on the basis of signal copies that have
sufficient signal energy or SNR above a minimum permissible value [139].
The order of diversity is decided by M transmitting antennas in M transmitting
and one receiving antenna system. In case of M transmitting antennas and N receiving
antennas, the diversity achieved is of order M*N.

4.2.3.2. Receive diversity
In contrast to transmit diversity, as shown in figure 4.4, the receive diversity employs
one transmitting antenna and a large number of receiving antennas. The apparent
drawback of transmit diversity includes extra consumption of bandwidth due to
duplication of the information signal to be sent.

Figure 4.4: Receive Diversity having multiple antennas at receiver side

This problem is avoided in receive diversity scheme with increased antenna
complexity. The signal from transmitting antenna travels through multipath to reach
to the multiple receiving antennas related to the receiving section. These received
copies of signal experience independent fading through independent paths in the
channel which are further processed according to various receive diversity
mechanisms. The recovered signal, thus, though receive diversity mechanism,
demonstrates superior signal strength.

73

4.3. Combining Mechanisms
In this section, various signal combining mechanisms related to receive diversity will
be discussed in brief.
The idea of diversity is to combine several copies of the transmitted signal,
which undergo independent fading, in order to increase the overall received power.
Various combining methods for diversity scheme include different mechanisms of
treating multiple copies of received signal for enhancing the net received signal
strength at the receiver end.
In communication systems, for a slowly flat fading channel, the equivalent
low pass of the received signal of the branch i can be written as:
( ) . ( ) ( )
j
i i i
r t Ae s t n t
u
= + i =0, 2,.., M-1 .(4.4.1)
where, s (t) is the equivalent low pass of the signal

j
i
Ae
u
is the fading attenuation of branch i
is the AWGN channel noise ( )
i
n t
Here, for M branches in the transmission channel, M replicas of the transmitted signal
are obtained,
1 2 1
( ) [ ( ), ( ),........, ( )]
i
r t r t r t r t

=
M
...(4.4.2)
These M replicas of the transmitted signal are combined together depending
on various mechanisms discussed in the next sub-sections for achieving the
satisfactory SNR at the receiver end.

4.3.1. Selection combining
In this method, as shown in figure 4.5, a scanner switch continuously scans the status
of received signal replicas above a pre-decided threshold level of signal. The signal
replica with highest signal strength is connected with receiver for further processing.
74

The scanning process is a continuous process in order to get highest signal strength as
soon as the signal strength related to already connected branch goes below the pre-
decided threshold level.
For M branches or multiple paths in the transmission channel, with signal to
noise ratio achieved on each branch is
i
(i=1, 2 M). Further, it is assumed that the
received signal on each branch is independent and Rayleigh distributed with mean
power of 2
2
o . This type of diversity is microscopic in nature and is designed to
combat small scale fading.
The pdf of
i
is given by:
P (
i
) = (1/
0
) exp (-
i
/
0
) ... (4.4.3)
where i =1, 2...M &
0
0
The combining method of selection diversity is by picking the best branch of
the set of received branches by comparing each one with every other branch amongst
available SNRs of various branches [141].

Figure 4.5: Mechanism of Selection combining

75

The performance of diversity systems is usually measure by comparing the CDFs of
various combining methods.
Diversity of order 2 is generally preferred for the best trade-offs. The main
handicap of selection diversity is that the signals must be monitored at a rate faster
than that of the fading process if the largest of them all is to be selected [141, 113].

4.3.2. Maximal ratio combining (MRC)
In MRC mechanism, all the branches are processed simultaneously by weighting with
a gain factor proportional to its own SNR. Co-phasing and summing is performed for
adding up the weighted branch signals in phase.

Figure 4.6: Maximal Ratio Combining

The gain associated with the branch is decided by the SNR of the
corresponding branch, i.e. . The MRC scheme requires that the signals be
added up after bringing them to the same phase.
th
i
i
( / )
i
l S N =
If is the signal envelope, in each branch, then the combined signal envelope
is given as [141],
i
a
76


1
M
i i
i
a a
=
= l

....(4.4.6)
Assuming that the noise components in the channel are independent and
identically distributed in each branch, total noise power is

2
0
1
M
t
i
N N l
=
=
i
... (4.4.7)
The resulting SNR is thus given by

2
2
1 2
2
0 0
1
M
i i
i b b
M
i
i
a l
E E
a
N N
l

=
=
| |
|
| | | |
\
= =
| |
| | \ . \ .
|
\ .

.
(4.4.8)
Utilizing the Cauchy-Schwarz Inequality,

(
2
0
b
i
E
a
N
)

| |
s
|
\ .

..(4.4.9)
The equality in this case is obtained when *
i
l k a
i
= , where k being some
constant based on various other parameters. The maximum value of the output SNR
after MRC is given by

(
2
0
b
i
E
a
N
)

| |
=
|
\ .

(4.4.10)
i.e.
( )
2
0
b
i
E
a
N
i

| |
= =
|
\ .

(4.4.11)
(All summations between i=1, 2 ...M).
Thus it is noticed that the sum of the SNRs of the individual branches yields
the final SNR of the output. The main challenge in MRC combining technique is the
co-phasing mechanism related to the incoming branches after weighting them. The
expected value of the signal strength,
E [Signal Strength] = E [
i

] = M *
0
;
77

Therefore, the expected value increases linearly with the increase in M. The mean
value increases considerably as compared to the Selection combining [141, 113].

4.3.3. Equal gain combining (EGC)
The third type of combining method is the Equal Gain combining wherein the
gains s of the MRC scheme are all made equal to 1, for all i=1, 2 .M. Thus co-
phasing and summing is done on the branches which are received directly.
1
M
i
i
a a
=
=

The EGC is worse than MRC approximately by a factor of /4 0.8 [141].

4.4. Performance Evaluation of IDMA with MRC Diversity

The Transmitter and Receiver model with IDMA scheme has already been discussed
with their mathematical modeling.

4.4.1. IDMA Scheme with Maximal Ratio Receiver Combining (MRRC)
Diversity
The figure 4.7 demonstrates interleave-division multiple-access data based on
Maximal Ratio Receiver Combining (MRRC) diversity [113]. In this system, the
signal is transmitted through one antenna and is received with the help of two receive
antenna. Here, MRRC diversity technique is being employed due to its better
performance in comparison to other methods. In MRRC technique, the diversity
branches are weighted for maximum SNR. As shown in block diagram in figure 4.7,
after encoding and spreading the data is randomly interleaved and termed as chips.
Now, the chip signal is sent from the transmit antenna, which will propagate from
both the channel. Considering one transmits and two receive antenna diversity
78

channel 0 is denoted by and channel 1 is denoted by . These channels are
modeled by having magnitude and phase response.
0
h
1
h
So,
0
0 0
h =
i
e
u
o

1
1 1
h =
i
e
u
o (4.5.1)
Noise can be added at both the receiver. The resulting received signal is
0 0 k 0
R h x n = +

1 1 k 1
R h x n = + (4.5.2)
where and represents the noise and interference at both the receiver separately.
0
n
1
n

Figure 4.7: IDMA with MRRC Receiver diversity

Now the receiver combining scheme for two branches MRRC can be written
as [113],
79


* *
0 0 1 1 K
X h R h R = +

(4.5.2)


Now this output of maximal ratio combiner can be fed to the detector for the proper
estimation of signal
K
X . This is already discussed in receiver section.

4.4.2. IDMA Scheme with Maximal Ratio Transmitter Combining (MRTC)
Diversity
In [113], the typical Alamoutis transmit diversity is duly presented, where two or
more transmit antennas are used to transmit the information while one receive antenna
is employed to receive the same information. Basically this type of diversity, also
known as Maximal Ratio Transmitter Combining (MRTC) Diversity, is the remedy of
the problems associated with receives diversity i.e. the size of receiver and power
requirement at the receiving side.
In figure 4.8, interleave-division multiple-access with transmit diversity of
order 2, has been demonstrated. The channel between transmit antenna Tx0 and
receive antenna is designated by h0 while channel between transmit antenna Tx1 and
the same receive antenna is h1. The channel can be modeled with the help of following
equation [113],


0
0 0 0 0
( ) ( )
j
h t h t T h e
u
o = + = =

1
1 1 1 1
( ) ( )
j
h t h t T h e
u
o = + = = (4.5.3)
From the transmitter, at a given symbol period, two signals are simultaneously
transmitted from the two transmitting antennas.
80



Figure 4.8: MRTC diversity having two transmitter and one receiver antenna

The encoding is performed both in space and time domain. The encoding,
however, may also be accomplished in space and frequency. In this method, instead of
utilizing two adjacent symbol periods, two adjacent carriers have to be employed. The
received signal can be written as
0 0 0 1 1
( ) r r t h s h s n
0
= = + +

* *
1 0 1
( ) ( ) r t r t T h s h s n
1 0 1
= + = + + . (4.5.4)
In the combiner block, the signal is combined and sent to IDMA detector,

*
0 0 0 1
s h r h r = +
*
1
*
1
*
1
n h
*
1
n h
... (4.5.5)
*
1 1 0 0
s h r h r =
Putting the value of (4.5.3) and (4.5.4) in to (4.5.5),
2 2 *
0 0 1 0 0 0 1
( ) s s h n o o = + +
(4.5.6)
2 2 *
1 0 1 1 0 1 0
( ) s s h n o o = + +
81

The results combined in the detector are equivalent to two-branch MRRC
technique [113]. The only difference is accorded in the phase rotations on the noise
components which do not degrade the effective SNR. Therefore, the resulting
diversity order from the new two-branch transmit diversity scheme with one receiver
is equivalent to that of two-branch MRRC.

4.5. Simulation Results
In this section, simulation results are being presented to evaluate the performance of
IDMA with and diversity employing various parameters such as data lengths,
iterations, and many more. The interleavers used for the comparative analysis
includes Random Interleaver (RI), Master Random Interleaver (MRI), and Tree Based
Interleaver (TBI). The unit of bit energy to noise ( ) may be termed in dB
wherever applicable during the simulation of respective IDMA systems.
0
/
b
E N

4.5.1 Simulation Results of IDMA Scheme using MRRC Diversity
For performance evaluation of IDMA scheme with MRC diversity, it is assumed that
there is no adjacent-channel and co-channel interference. Also, the mobility of the
user in the cell is assumed to be negligible. For initial instance, the performance of
IDMA scheme will be evaluated with receiver diversity and later, the performance of
IDMA scheme with transmit diversity will be observed. For the purpose of
simulation, one transmitter and two receivers architecture is employed, initially.
In the figure 4.7, the IDMA scheme is shown with receiver diversity with one
transmitter and two receiver architecture. Figure 4.9 demonstrates the performance of
IDMA system with and without diversity with random interleaver, when data length
k=2048, spread length sl=16 is taken. At the receiver side, number of iteration taken
82

are 10 which is necessary to produce correct decoded results. The number of block is
taken = 2000. In the figure, it is demonstrated that the IDMA scheme performs better
with diversity implementation due to MRRC mechanism.
While choosing the iteration count at the receiver, it is necessary to decide
optimum count of iterations in turbo processing of data at the receiver. In figure 4.10,
the performance of IDMA System is shown with variation in iteration count and it is
demonstrated that with increment in iteration count the BER performance of system
improves considerably. However, a trade-off has to be observed for iteration count at
the receiver in order to keep the computational complexity in optimum limit. Also, it
can be easily observed form [103] that the system performance does not improve in
proportion to the complexity incurring due to increment in count of iterations, after
optimum count of iterations, on receiver side.

1 2 3 4 5 6 7 8
10
-5
10
-4
10
-3
10
-2
10
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e


without diversity
with MRRC diversity

Figure 4.9: Performance of IDMA system with and without MRRC diversity
83


1 2 3 4 5 6 7 8
10
-5
10
-4
10
-3
10
-2
10
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e


iteration=1
iteration=5
iteration=10

Figure 4.10: Performance of IDMA having Random Interlever with MRRC diversity
technque at various iterations count with datalength=1024, spreadlength=16.

In figure 4.11, the parameter of data length is varied with k=512, 1024, 2048.
The simulation is performed with one transmitter two receiver diversity systems.
From the simulation, it is demonstrated that with increment in data length, the BER
performance of the system is improved due to increment in available count of
interleavers as explained in section 3.4. However, interleavers required for specific
user remains, still, negligible in comparison to total available interleavers. In the
simulation other parameter taken are block length=2000, spread length=16, and no. of
iterations=15.
In figure 4.12, the performance of IDMA scheme with tree based interleaver
(TBI) is demostrated with and without diversity mechenism. Data length taken for
simulation is 1024 bits with 2000 bits as block length and 16 as spread length. The
iteration for turbo processing is taken to be 15. With the help of simulations, it can be
84

concluded that IDMA is performing better when it is implemented with MRRC
diversity technique.
In figure 4.13, the IDMA scheme with variation in interleaver is demonstrated.
In the simulation no forward error correction (FEC) is used. The datalength and
spreadlength are taken to be 1024 bits and 16 respectively. The iterations at the
receiver are taken as 10. The simulation results demontrates the similar pefrmance of
tree based interleaver in comparison to other interleavers. However, it it worth to be
noted that tree based interleaver employs optimum bandwidth and coputational
complexity when compared with similar parameters of other inteleavers.
In figure 4.14, 4.15, and 4.16, the performance of IDMA scheme with random
interleaver (RI), master random interleaver (MRI), and tree based interleaver (TBI)
individually with variation in datalengths i.e . 124 bits, 512 bits, and 1024 bits.
1 2 3 4 5 6 7 8
10
-5
10
-4
10
-3
10
-2
10
-1
Eb/No(dB)
B
i
t

E
r
r
o
r

R
a
t
e


datalength 512 without diversity
datalength 512 with diversity
datalength 1024 without diversity
datalength 1024 with diversity
data lenghth 2048 without diversity
datalength 2048 with diversity
Figure 4.11: Performance comparison at different data length with Random Interleaver with
MRRC Diversity

85

1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6
10
-3
10
-2
10
-1
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e
y g
k=1024
k=1024
without diversity
with diversity

Figure 4.12: Performance of IDMA using tree based interleaver with MRRC Diversity


1 1.5 2 2.5 3 3.5 4 4.5
10
-6
10
-5
10
-4
10
-3
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e


with Random Interleaver
with Master Random Interleaver
with Tree Based Interleaver

Figure 4.13: Simulation of Uncoded IDMA with variation in interleaver

86

1 2 3 4 5 6 7 8 9 10
10
-4
10
-3
10
-2
10
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e


Without diversity datalen=124
Without diversity datalen=512
Without diversity datalen=1024
With diversity datalen=124
With diversity datalen=512
With diversity datalen=1024

Figure 4.14: Performance of Random Interlever with 1Transmitter and 2 Receive Antenna,
MRRC diversity Technique

1 2 3 4 5 6 7 8
10
-4
10
-3
10
-2
10
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e

without diversity datalen=1024
without diversity datalen=512
without diversity datalen=124
with diversity datalen=124
with diversity datalen=512
with diversity datalen=1024

Figure 4.15: Performane of master random interleaver with 1Transmitter and 2 Receive
Antenna, MRRC diversity Technique


87

1 2 3 4 5 6 7 8 9 10
10
-4
10
-3
10
-2
10
-1
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e


With out diversity datalen=124
Without diversity datalen=512
Without diversity datalen=1024
With diversity datalen=512
With diversity datalen=1024
With diversity datalen=124

Figure 4.16: Performance of Tree based Interleaver with 1Transmitter and 2 Receive
Antenna, MRRC diversity Technique


In each of the simulations, the one transmitter and two receiver architecture
has been used. The spread length and iteration count are chosen to be 16 and 15
respectively. The performance of the IDMA system confirms the superiority of
diversity techniques in the implementations.
In figure 4.17, the perfromance of IDMA scheme is observed with rate
convolutional forward error correction (FEC) coding using tree based interleaver with
1024 bits of datalength. The simulation results demonstartes that with FEC coding
and MRRC diversity, the results are outperforming the other simulations. The results
without diversity and without FEC coding, are far inferior to that with FEC coding
and MRRC diversity scheme.
88

1 2 3 4 5 6 7 8
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e


MRC diversity with FEC coding
MRC diversity without coding
no diversity no coding
no diversity with FEC coding

Figure 4.17: Perfromance of IDMA scheme with FEC coding with tree based interleaver and
datalength=1024 bits.


1 2 3 4 5 6 7 8
10
-7
10
-6
10
-5
10
-4
10
-3
10
-2
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e


1 tx 8 rx
1 tx 6 rx
1 tx 4 rx
1 tx 2 rx
Figure 4.18: IDMA scheme with variation in receiver count using MRRC diversity
89

In space diversity, if the number of receivers or transmitters are increased in
receive or transmit diversity respectively, the BER performance of the systems
improves drastically. However, with increment in receiver and transmitter count, the
cost of the system increases rapidly. The transmit diversity has additional drawback of
consmption of extra nadwidth in the channel. The results are verified with simulations
presented in figure 4.18.
In the figure 4.18, for simulation purpose, tree based interleaver has been
employed. The simulation results demonstrates the superior BER performance of the
system with higher receiver count to that of with lower receiver count.

4.5.2 Simulation Results of IDMA Scheme using MRTC Diversity
As reportd in the literature [113], reciever diversity techinque is classical apprach to
achieve diversity but due to receiver size and power requirement transmit diversity
technique is overcoming the receive diversity.However, as already stated, transmit
diversity suffers woth drawback of employing more bandwidth. Still, it demonstrates
similar BER performance to that of for receiver diversity.
In this section, transmit diversity is discussed duly implemented with IDMA
systems denoted by maximal ratio transmit combining (MRTC) diversity scheme. The
basic block diagram is shown in figure 4.8. In the figure, the transmit diversity with
two transmit and one receive antenna system has been shown. The assumptions made
for simulation of receive diversity are still valid.
90

1 2 3 4 5 6 7 8
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e
y g g y
Random interleaver without diversity
Random Interleaver with transmit diversity

Figure 4.19: Performance of IDMA system using MRTC diversity with random interleaver

In figure 4.19, performance of IDMA is evaluated using transmit diversity
employing random interleaver for the purpose of simulation. Other parameters chosen
for the simulation includes 2048 bit as datalength, spread length to be 16, no. of
iteration to be 15, and block length to be 200. The simulation results demonstrates the
superior performance of transmit diversity over the system without diversity. In the
simultions, no forward error correction (FEC) has been employed.
The figure 4.20 is demonstrating the performance of IDMA with MRTC
diversity employing variation in datalengths. The datalength chosen for simulation are
512 ,1024 2048 bits. Other parameters are taken to be block length to be 2000,
iteration count to be 15, and spread length to be 16 for the analysis.
91

1 2 3 4 5 6 7 8
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e
y g g y
k=512
k=1024
k=512
k=1024
k=2048

Figure 4.20: Performance of IDMA at different data lengths using random interleaver with
MRTC diversity Scheme

In figure 4.21 shows the simulation results of IDMA using tree based
interleaver. The simulation is performed in uncoded environment. The simulation
results demonstrate the similar performance of all the interleavers for similar
datalengths. However, higher datalengths are prone to produce better BER
performance in comparison to that of with lower datalengths. The data lengths are
taken to be 256, 512, and 1024 for simulation purpose. The other parameters are
chosen to be spread length 16, block length 2000, and iteration count 15 for the
purpose of simulation.
In the silulation results shown in figure 4.22, rate convolutional codes are
employed as FEC codes. The other silulation parameters are similar to that employed
for figure 4.21. The conclusions of this simlations are also similar to that of previous
simulation.

92

1 2 3 4 5 6 7 8
10
-7
10
-6
10
-5
10
-4
10
-3
10
2
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e


TBI uncoded with datalength 512 bit
RI uncoded with datalength 256 bit
TBI uncoded with datalength 256 bit
RI uncoded with datalength 512 bit
TBI uncoded with datalength 1024 bit
RI uncoded with datalength 1024 bit

Figure 4.21: Uncoded IDMA scheme with variation in datalength for RI & TBI


1 1.5 2 2.5 3 3.5 4 4.5 5
10
-7
10
-6
10
-5
10
-4
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e


TBI coded with datalength 1024 bit
RI coded with datalength 1024 bit
TBI coded with datalength256 bit
TBI coded with datalength 512 bit
RI coded with datalength512 bit
RI coded with datalength256 bit


Figure 4.22: Rate Conlutionaly coded IDMA scheme with variation in datalength for RI &
TBI

93

1 2 3 4 5 6 7 8
10
-7
10
-6
10
-5
10
-4
10
-3
10
2
Eb/No
B
i
t

E
r
r
o
r

R
a
t
e


TBI uncoded with datalength 1024 bit
RI uncoded with datalength 1024 bit
TBI coded with datalength 1024 bit
RI coded with datalength 1024 bit

Figure 4.23: Rate Conlutionaly coded IDMA scheme with uncoded IDMA scheme


In figure 4.23, the performance of rate convolutional coded IDMA scheme
is compared to that of uncoded IDMA scheme with 1024 bits of datalengths. The
other simulation parameters are similar to that of previous simulations.
The perfornace demonstrated in the simulatons, confirm the superior
perfromance of IDMA systems with transmit diversity .

4.6 Conclusions
It is well established that diversity is the solution of fading without the increment in
complexity and extra requirement of bandwidth in the system. In this chapter, various
types of diversity schemes have been discussed, i.e. receive diversity and transmit
diversity. In addition to it, various combining schemes has also been presented,
including selection combining, maximal ratio combining and equal gain combining.
All the necessary mathematics of each diversity and combining schemes has also been
94

95

discussed. Various simulations presented confirm the superior performance of IDMA
scheme with diversity systems.
















CHAPTER 5
Correlation Analysis and FPGA Implementation of
Interleavers



5.1. Introduction
The importance of interleavers has been established as integral component in IDMA
systems. For proper functioning of the IDMA systems, it becomes important to
consider their design criterion with utmost care. During designing of user-specific
interleavers, emphasis has to be given to the non-resemblance of the interleavers with
each other because they are being used for the purpose of user separation in IDMA
systems. Also, attention has to be paid for efficient design of these interleavers so that
they should easy to implement with least hardware with maximum speed of operation.
In this chapter, the correlation analysis of RI, MRI, and TBI has been explored in
order to compare their resemblance with other interleavers in their respective
categories. Also, all the above interleavers have been implemented on field-
programmable gate-array (FPGA) for comparison of hardware and timing constraints.

5.2. Motivation
In CDMA systems, PN sequences are employed for the user separation while
interleavers are used with them for the purpose of providing better coding gain. In
case of IDMA systems, the interleavers are required for the sake of user separation
[71, 76]. Since the separation of users is achieved by user-specific interleavers, an
96

interleaver design criterion includes that every two interleavers out of a set of
interleavers must avoid collision up to maximum extent. With increment in
collision amongst user-specific interleavers, the BER performance of system goes for
dips. For the reason, these user-specific interleavers must be orthogonal in nature in
order to avoid the collision amongst them.
Since the separation of users is achieved by interleavers, an obvious
interleaver design criterion is that every two interleavers out of a set of interleavers
collide as little as possible. The goal in this section is to define correlation among
interleavers for IDMA in order to measure the level of collision among interleavers.
Unlike in turbo coding/decoding mechanism, where the task of a any interleaver is to
decorrelate different sequences of bits [95, 53], in IDMA systems, the interleavers in a
set of interleavers should need to decorrelate user-specific chip sequences. The
correlation between interleavers should measure how strongly signals from other
users may affect the decoding process of a specific user. Hence, the additive noise
should not play a role in the correlation of interleavers. Throughout entire section, the
noiseless IDMA system is considered for analysis.
In addition to it, the implementation of technology with comparative minimum
hardware along with timing constraints is also an important issue for researchers. In
[145, 146], the Very High Speed Integrated Circuit Hardware Description Language
(VHDL) and FPGA implementation of convolutional interleaver/ deinterleaver have
been presented. In addition to it, many other researchers have contributed towards
hardware implementation of orthogonal and non-orthogonal interleavers. However,
significant work has to be, still, done in the field of hardware implementation of
orthogonal interleavers. In this chapter, the FPGA implementation of RI, MRI, and
97

TBI mechanisms have been duly presented along with their comparison of hardware
requirements and timing constraints.

5.3. Design Criteria for Interleavers in IDMA scheme

Random interleavers for IDMA are needed to satisfy two design criteria [76, 91]
which are as stated below.
1) They should be easy to specify and generate, i.e., the transmitter and receiver can
send a small number of bits between each other in order to agree upon an interleaver,
and then generate it.
2) The cross correlation between them should be minimum for avoiding the collision.
The collision among interleavers is interpreted in the form of the uncorrelation among
the interleavers.
If the interleavers are not randomly generated, the system performance
degrades considerably and the MUD is unable to resolve MAI problem at the receiver
resulting in higher values of Bit Error Ratio (BER). On the other hand if the
interleaving patterns are generated more and more random, the MUD resolves the
MAI problem more quickly and better values of BER are obtained for the same
parameters.

5.4. Correlation in Interleavers

The correlation of two interleavers is most important criterion in the IDMA system.
The correlation of any two sequences lies between -1 to +1. If the correlation between
two sequences is minimum or zero, it implies that these sequences do not match to
each other. For ideal orthogonality condition, the value of cross-correlation between
two sequences should be zero.
98

Let Y1 and Y2 be two variables with joint pdf p(y1, y2), then their joint
moment of order (k, n) is defined as,

1 2 1 2 1 2 1 2
[ ] ( , )
k n k n
E Y Y y y p y y d y d y


=
} }
... (5.3.1)

When k=n=1, the joint moment in such a case is known as correlation. Hence,
correlation is defined as

1 2 1 2 1 2 1 2
[ ] ( , ) E Y Y y y p y y d y d y


=
} }
. (5.3.2)

Definition of Correlation and Orthogonal Interleavers [91]:

Definition 1: Let
i
| and
j
| be two interleavers and x and y be two words
respectively. Further, defining the correlation C (
i
| , x,
j
| , y) between
i
| and
j
| with
respect to the words w and v as the scalar product between
i
| (f(x)) and
j
| (f(y)):
C (
i
| , x,
j
| , y) = {
i
| (f(w)),
j
| (f(v))} .. (5.3.3)
Definition 2: Two interleavers
i
| and,
j
| (where
i
|
j
| ) are called orthogonal, if
for any two words x and y,
C (
i
| , x,
j
| , y) = {
i
| (f(x)),
j
| (f(y))} = 0....(5.3.4)
It is easy to verify that if a set of mutually orthogonal interleavers is used in
the IDMA system, then they do not collide to each other. In this sense, zero-
correlation (or orthogonality) implies no collision among interleavers.
Evaluation of the correlation between two interleavers with respect to every
possible pair of two words is computationally complex. It is due to reason of various
possibilities of combination of every word.

5.5. Correlation Analysis of Interleavers
For performing the correlation analysis on considered interleavers including, RI, MRI,
and TBI, the noiseless IDMA system is taken into consideration. Further, for
99

calculation of cross-correlation amongst interleaver in their respective categories, the
parameters utilized in the simulation are kept fixed for all the interleaving
mechanisms. All the simulations are performed on MATLAB 7.0 environment. The
other parameters opted for simulation includes user count to eight with data of eight
chips/ user.
For calculation of cross-correlation, for fixed user count, the cross-correlation
of all the user-specific interleaving sequences is calculated with each other. For
example, in case of five users, the cross-correlation of first interleaving sequence,
with the other entire four users is calculated. Later, the cross-correlation of second
interleaving sequence with all the other sequences is, again, calculated. This process
is repeated for all the user-specific interleaving sequences. The calculated cross-
correlations related to all the users are further added in order to calculate overall
cross-correlation. The resultant user-specific cross-correlation represents the
similarity of data-spreaded user-specific interleaving sequences amongst each other.
The higher value of resultant cross-correlation represents the sign of lower de-
correlation amongst user-specific interleavers. It basically represents the increment of
multiple access interference (MAI) in the system. As a desired parameter, the
resultant user-specific cross-correlation should be minimum up to extent of maximum
possibility.
The cross correlation at a particular instant of time, for a particular count of
user, will be decided by user-count in cell and independently generated user-specific
interleavers. For example, the cross correlation between first and second user, in case
of 25 users and 100 user, as shown in figure 5.2 and 5.3, is 1.76 and 10.0 respectively.
The flowchart for the mechanism adopted for analysis is shown in figure 5.1.
100

The results of resultant user-specific cross-correlation are demonstrated for 25
users and 100 users in a cell at a time, however, the simulations were performed for
other user count including 10, 50, and 80 also. The first result of resultant user-
specific cross-correlation with 25 users is presented for RI, MRI, and TBI in figure
5.2. Another result of resultant user-specific cross-correlation with 100 users is
demonstrated in figure 5.3. The interleavers were generated using random, master
random or tree based algorithms respectively. As the user count increases, the overall
cross-correlation amongst users increases.

Figure 5.1: Mechanism for calculation of resultant user-specific cross-correlation amongst
users

From the figure 5.2 and 5.3, it is observed that the Resultant user-specific
cross-correlation increases with user count. In happens due in loss of resemblingness
amongst user-specific interleaving sequences. Therefore, as the user count increases,
101

the cross-correlation between users also increases which further leads to increment in
bit error rate (BER) of the system. The results observed from figure 5.2 and 5.3, apart
from other results, are listed in table 5.1.

Random Interleaver with 25 users
0
0.5
1
1.5
2
2.5
3
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
User Number
C
r
o
s
s

C
o
r
r
e
l
a
t
i
o
n

Master Random Interleaver with 25 Users
-0.5
0
0.5
1
1.5
2
1 3 5 7 9 11 13 15 17 19 21 23 25 27
User Number
C
r
o
s
s

C
o
r
r
e
l
a
t
i
o
n

Tree Based Interleaver with 25 Users
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1 3 5 7 9 11 13 15 17 19 21 23 25
User Number
C
r
o
s
s

C
o
o
r
e
l
a
t
i
o
n

Figure 5.2: Resultant user-specific cross-correlation for 25 users with RI, MRI, and TBI



102

Master Random Interleaver with 100 Users
-10
-5
0
5
10
15
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99
User Number
C
r
o
s
s

C
o
r
r
e
l
a
t
i
o
n

Random Interlever 100 Users
-15
-10
-5
0
5
10
15
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99
User Number
C
r
o
s
s

C
o
r
r
e
a
l
t
i
o
n

Tree Based Interleaver with 100 Users
-10
-5
0
5
10
15
1 8 15 22 29 36 43 50 57 64 71 78 85 92 99
User Number
C
r
o
s
s

C
o
r
r
e
l
a
t
i
o
n

Figure 5.3: Resultant user-specific cross-correlation for 100 users with RI, MRI, and TBI




103

Table 5.1

Peak Resultant User-Specific Cross-Correlation of RI, MRI, and TBI

Peak Resultant User-Specific Cross-Correlation Sr.
No.
User
Count Random
Interleaver (RI)
Master Random
Interleaver (MRI)
Tree Based
Interleaver (TBI)
1. 10 1.22 1.18 1.22
2. 25 2.76 1.76 1.52
3. 50 4.06 4.35 4.05
4. 80 6.18 6.25 6.2
5. 100 11.2 12.2 10.0

In figure 5.4, the graphical representation of results listed in table 5.1, are
demonstrated. From the results, it is observed that for lower user count, the resultant
cross-correlation related to user-specific interleaving sequences, is similar for all the
interleavers with lower values. So, at this stage the BER performance of IDMA
systems with all the interleavers is nearly same. With increment in user count the
value of resultant cross-correlation related to user-specific interleaving sequences,
also increases. The increment in user-specific cross-correlation results in increment of
multiple access interference (MAI) in the system, which ultimately increases the BER
of the systems. However, the value of resultant user-specific cross-correlation, for
random interleaver, is observed to be at lowest level, which confirms the better BER
performance of random interleavers at higher user count in comparison to other user-
specific interleavers.
104


Figure 5.4: Graphical view of resultant user-specific cross-correlation with RI, MRI, and TBI


The user-specific cross-correlations of tree based interleaver and master
random interleaver is also observed to be at similar level in case of higher user count.
Due to this reason, the BER performance of all the discussed interleavers is similar
for higher user count.

5.6. Interleaving Mechanisms in IDMA Scheme

As already stated in chapter 2, user-specific interleavers are employed as only means
for user separation in IDMA scheme. These interleavers are selected from the
interleaving combinations available from total count of chips. If the selection of user-
specific interleavers is performed randomly for all the users, the selected interleavers
are called random interleavers. These randomly selected interleavers, at transmitter
105

end, are required to be sent to receiver end for proper decoding of user-specific data
which needs high bandwidth and memory requirement at various locations.
In [104], the proposed interleaver omits both of stated problems by randomly
selecting only one interleaver known as master random interleaver (MRI) or power
interleaver. Other user-specific interleavers are further generated by multiple re-
interleaving of master interleaver depending on user count. As an example, for
generation of interleaver for 30
th
user, the MRI has to re-interleave for further 29
th

times. Also during transmission, only MRI and user count has to be sent to the
receiver end for calculation of user-specific interleaving patterns. However, MRI
produces the problem of computational complexity in the receiver during turbo
processing [62].
In [147], the proposed tree based interleaver (TBI) solves all the problem of
concerns optimally. In TBI, two master interleavers are selected randomly out of
available random interleavers. The other user-specific interleavers are computed with
the help of both of the master interleavers and user count as shown in figure 3.1 in
section 3.3. However, during the transmission, two master interleavers and user count
along with data has to be transmitted to the receivers. It is here worth to mention that,
in random interleaver (RI), all the user-specific interleavers were to be sent to receiver
while in case of MRI, one master interleaver and the user count was sent to the
receiver along with the data related to all the users. The data formats for RI, MRI,
and TBI are demonstrated in figure 3.6 in section 3.4.
The interleavers used for user separation are allotted at the transmitter end and
either the interleaver itself of some related parameters have to be sent along with data.
During the communication process, due to dynamic nature of mechanism related to
user-specific interleaver allocation, it becomes mandatory that the related mechanisms
106

should be efficient and least complex in terms of hardware requirement and timing
constraints.
In next section, the functional diagrams of RI, MRI, and TBI will be discussed
with their implementations on field-programmable gate-array (FPGA) on XILINX
platform. The implementations have been performed for 8 users with data of 8 chips/
user only. Also, the performance comparisons for all the interleavers, on the
parameters of hardware requirements and timing constraints, have been duly
presented.
From the discussion, it becomes evident that in master random interleaver, the
computational complexity increases with increment in data length and user count
while in case of random interleaver the computational complexity is at its minimum
level while there is high requirement of registers and flip-flops for storing the
interleavers.

5.6.1. Random Interleaving (RI) Mechanism

The random interleavers are randomly selected interleavers depending on user count,
from the available combination-patterns of total chips. Figure 5.5 presents the
functional block diagram of random interleaver. During the generation of random
interleavers, it is ensured that no two positions for total chip lengths are repeated in
any case. The block diagram is divided in two parts including random position setter
(RPS) and data transfer sections. In the RPS section, the randomly selected two
positions are generated with assurance that these positions have not been used
previously during the current interleaver generation. However, in data transfer
section, the user-specific data is transferred in the generated interleaving positions. In
the RPS section, the random incrementor-1 (RI1) and random incrementor-2 (RI2) are
107

used to generate a random position count. These randomly selected positions are used
for swapping the data related to user, in the data transfer section. The shown block
diagram is valid for input of two or more than two bits of data.


Figure 5.5: Block Diagram of Random Interleaving Mechanism

Initially, all the swapping position multiplexers (MUXs) and other blocks are
reset in order to ensure that no initial value is left stored with them. The data length
and user data is fed into appropriate block as in the shown diagram. During the
initiation, the fist and second random incrementor blocks are loaded with certain
random positions. The random position with RI1 block is forwarded to first swapping
position MUXs bank with initial clock which is further added with the random
position inside RI2 block. The modulus counting block ensures that the output of RI2
108

block is within limits of available datalength of chips. Both the swapping positions are
then compared to avoid the repetition of the positions. In case of repetition of
positions, the signal is fed to RI1 for re-generation of another random position.
Further, these non-repeated swapping positions are provided to swapper bank block
which swaps the user-specific data available at its other input. In addition to it, the
count comparator block compares the input of position counter with half of the
provided data length to confirm that the position selected are not going beyond of
total data length which is in fact in form of chips.
5.6.2. Master Random Interleaving Mechanism

In case of master random interleaver (MRI), the user-specific interleavers are
generated with the help of randomly selected master interleaver and user count [104],
at transmitter end. The random position setter (RPS) block, in figure 5.6, generates the
interleaving positions for master random interleaver. Later, with the help of user count
and master random interleaver, the other user-specific interleavers are decided.
The random position setter (RPS) block functions in a similar manner as was
recognized in figure 5.5. The master random sequences, generated by RPS block, are
provided to temporary register for their storage. After getting, the entire master
interleaving sequence, the temporary register block disables the RPS block so that no
further generation of master sequence may take place. Further, the temporary register
provides the stored master interleaving positions to the loop-count with user (LCU)
block. This block decides the other user-specific interleaving sequences by looping of
first generated master random sequence as per master random interleaving algorithm,
depending on user count.
The algorithm for master random interleaver generation involves one master
interleaver i.e. t which is allotted to first user as user-specific interleaver. However,
109

for the second user, the interleaving sequence will be calculated by re-interleaving the
master interleaver t i.e. ( ) t t
))
.Similarly, for third user, the interleaving sequence will
be calculated from ( ( t t t . For the others users, the user-specific interleavers can be
calculated in similar continuing manner [104]. Therefore, the complexity of loops
increases with increment in data length and user count.

Figure 5.6: Block Diagram of Master Random Interleaver (MRI)

The calculated user-specific interleaving sequences are, further, provided to
swapper bank which swaps the positions of the interleaving sequence. The user-
specific data is, now, adjusted according to generated interleaving sequences. The
temporary register provides the interleaving positions for further reference.
5.6.3. Tree Based Interleaving Mechanism
The block diagram demonstrating all the inbuilt major integral blocks is shown in
figure 5.7. In case of tree based interleaving mechanism [147], the RPS block is
required to generate two master interleaving sequences which are stored in temporary
110

register. After receiving of both the master interleaving sequences, the temporary
register inhibits the RPS block for further generation of interleaving sequences.

Figure 5.7: Block Diagram of Tree based Interleaving mechanism

Further, the user specific interleaving sequences are generated based on tree algorithm
in LCU block. The LCU block loops both the master interleaving sequences with each
other according to tree algorithm described in section 3.3. In fact the LCU blocks
related to master random interleaver and tree based interleaver generation mechanism,
are similar except the algorithms accompanied with them.
In the next section, the FPGA implementation results of RI, MRI, and TBI will be
discussed. The implementation has been performed on FPGA in XILINX
environments for 8 users with 8 chips/ user.


111

5.7. Performance Comparison of Interleavers on FPGA
implementation
All the interleavers including RI, MRI, and TBI have been implemented on FPGA in
XILINX environment. The summary of the results related to FPGA implementations
are explained in next sub-sections.

5.7.1. Summary of hardware
In this section, the comparison of RI, MRI, and TBI is performed based on hardware
requirements for generation of user-specific interleavers at transmitter side. For the
comparison purpose, the parameters of comparison are selected to be D-flip flops,
comparator, adder/subtractor and multiplexers.
According to table 5.2, it is evident that the requirement of hardware
components for tree based interleaver is at minimum level in comparison to those for
others. Master random interleaver requires highest number of hardware components,
at transmitter end, due to number of user dependent loops involved for generation of
user-specific interleavers. While in case of RI generation mechanism, hardware
requirement does not increase significantly with increment in user count.
Table 5.2. Summary of hardware
Random
Interleaver (RI)
Master Random
Interleaver (MRI)
Tree Based
Interleaver (TBI)
D Flip-Flop 832 3520 104
Adder/Subtractor 272 1152 32
Comparator 128 576 16
MUX 576 2592 72


112

5.7.2. Final Register report
In final register report, the parameters for comparison are the registers and flip flops
required to prepare the interleavers. The final register comparison is shown in table
5.3. The requirement of registers and flip-flops for MRI is at the highest level in
comparison to that need for RI and TBI.
Table 5.3. Final Register report
Random Interleaver
(RI)
Master Random
Interleaver (MRI)
Tree Based Interleaver
(TBI)
Register 850 3528 108
Flip-Flop 850 3528 108

The tree based interleaver requires least amount of components in final
register report due requirement of lesser components for the storage of intermediate
interleaving sequences occurring during the calculation of user-specific interleavers.
In case of RI, the LCU block is not at all needed due to any need of looping but it
requires considerable amount of storage elements in RPS block to store all the user-
specific interleavers at transmitter end.
5.7.3. Device utilization Report
In table 5.4, the summary of device utilization is presented for RI, MRI, and TBI
interleavers. Here, the parameters for comparison are chosen to be count of slices,
look-up tables (LUTs), and inputs/ outputs (I/Os) ports, required for building these
interleavers. The look-up tables are basically an array or associative array, of data,
often used to replace a runtime computation with a simpler array indexing operation.
Due to look-up tables, the savings in terms of processing time can be significant, since
113

retrieving a value from memory is often faster than undergoing an 'expensive'
computation or input/output operation [4].
Table 5.4. Device utilization report
Random Interleaver
(RI)
Master Random
Interleaver (MRI)
Tree Based Interleaver
(TBI)
No. of slices 9394 48245 1198
No. of 4 input LUT 18207 94812 2328
No. of I/Os 130 130 17

In master random interleaver, more look-up tables are needed in order to store
intermediate calculations due to looping operation. However, in random interleaver,
significant count of look-up tables and I/Os port are required to store the generated
user-specific interleavers. In case of tree based interleaver, least amount of look-up
tables are needed for storage of data due to availability of only two master interleavers
and other associated intermediate results.
The requirement of I/Os is dependent on deployment of respective look-up
tables. Therefore, due to lesser requirement of look-up tables, the TBI needs least
amount of I/Os devices. From the above stated comparison, it is apparent that the tree
based interleaver requires least amount of devices while MRI requires the highest
number of components, for 8users with data of 8 chips/ user.

5.7.4. Timing Summary Report

In this comparison, the parameters opted are related to speed of operation for the
interleavers. This maximum frequency of operation demonstrates the capability of
user-specific interleaver for its operation. Parameters such as minimum arrival time
114

before clock and maximum output time required are important parameters
representing the performance of interleavers under high frequency environment.
When combined, all the above parameters demonstrate the overall timing behavior of
these interleavers.
Table 5.5. Timing summary of interleavers
Random
Interleaver
(RI)
Master Random
Interleaver
(MRI)
Tree Based
Interleaver
(TBI)
Minimum period 132.745 ns 264.074 ns 117.482 ns
Maximum
frequency
7.533 MHz 3.787 MHz 8.512 MHz
Minimum input
arrival time
before clock
34.131 ns 177.980 ns 33.680 ns
Maximum
output required
time after clock
6.788 ns 6.788 ns 6.788 ns

From table 5.5, it becomes clear that tree based interleaver can be operated at
maximum frequency with minimum time period requirements. The reason for its
fastest speed is inherent behind the requirement of least count of loops, and hardware
required for generation of user-specific interleavers. Therefore, it is found that the tree
based interleaver has the best timing behavior among all these interleavers and master
random interleavers require the largest time for its operation with similar clock
frequency.

5.8. Conclusions
In this chapter, the user-specific orthogonal interleavers are implemented on FPGA
and compared duly on various parameters, including hardware requirements and
timing constraints. The FPGA implementation was performed for eight users with
115

eight chips for each user. The concluding remarks for the work in this chapter are
given below for all the interleavers.
1- For master random interleaver (MRI), there is need of one master interleaving
pattern which further generates the other user-specific interleavers. Therefore,
with increment in user count, the computational complexity due to looping
operation increases. As the loop count increases, the memory requirement for
storage of intermediate results and associated hardware requirement also
increases. So, this interleaver has worst performance when compared with all
the other interleavers with increment in user count.
2- In case of random interleaver, there exist unique interleaving pattern for each
user. Depending on user count, the number of random interleaving sequence
increases, leading to requirement of higher memory requirement and
proportionally higher count of registers, flip-flops, and other hardware devices
for storage of sequences in RPS block. However, due to absence of looping
operations, the requirement of hardware is less than that required in case of
MRI. The Timing constraints of RI are also moderate in comparison to MRI
due similar reasons.
3- The tree based interleaver (TBI) needs two master interleaving sequences for
generation of other user-specific interleaving sequences. As the requirement of
looping operation in TBI generation mechanism, is quite less in comparison to
that of MRI, it inherits comparably lesser complexity. Also, only few
intermediate results, resulting in RPS block, are required to be stored when
compared to that for RI and MRI. Due to this reason, there is requirement of
least hardware for TBI generation leading to minimal timing requirement for
its operation. It is also worth mentioning that in user-specific interleaver
116

generation process; the number of looping operation affects the behavior of
generation mechanisms greatly due to large number of loops involved, leading
to large number of temporary variables for storage of intermediate data.
On the behalf of all the simulations performed on XILINX platform, the tree
based interleaver comes out to have best performance because it tree based
interleaving mechanism requires least count of hardware as well requires least on all
the comparisons.


















117

CHAPTER 6


CONCLUSIONS

Due to emerging demands on wireless communication, related to higher data rate
along with other recommendations of International Telecommunication Union (ITU),
it has become emergent to establish an optimal communication system with global
acceptability. The multiple access schemes are referred to as integral part of any
communication system in the world of communication engineering. In 1993,
Qualcomm Corporation in U.S.A., accommodated spread spectrum communication in
form of code-division multiple-access (CDMA) for meeting the demand of higher
data rates and other related facilities. However, due to emerging demands for more
user-compatible technology and few inherent problems in CDMA scheme, another
variant of CDMA scheme, known as interleave-division multiple-access is proposed
in [76, 80].
Interleave-Division Multiple-Access (IDMA) scheme is referred as the
spread spectrum multiple access technique in which user specific interleaver is
employed for individual user separation. Since then, numerous technical papers about
IDMA have been published in the literature. Moreover, at present, IDMA is being
considered to be a most promising candidate for the air interface of the future
generation in wireless communication systems.
In IDMA, orthogonal interleavers are the only means for user separation and,
hence, are referred as heart of IDMA scheme. The efficient selection of interleaver
along with optimum design methodology leads to satisfactory spectral efficiency in
IDMA systems. During initial run of IDMA scheme by Prof. Ping, random
118

interleavers (RI) were employed for user separation. Later, random interleavers
were replaced by master random interleavers (MRI) [104] for optimizing the
memory requirement raised due to storage of user specific interleavers at transmitter
and receiver ends. The presence of problems in random and master random
interleavers has led towards development of proposed optimum interleaver for
IDMA scheme. The work presented in this thesis is completely devoted to the
investigation of an optimum interleaver for IDMA scheme.
In this thesis, the performance and analysis of optimal user-specific interleaver
generation mechanism known as tree based interleaver has been presented [147].
Performance evaluation of this interleaver using analytical modeling and simulation
has been carried out in this work. During the performance analysis with due
simulations, it has been found that the computational complexity of proposed
interleaver is extremely less in comparison to that for master random interleaver while
it is marginally higher to that occurring in case of random interleaver. However, the
bandwidth and memory requirement of proposed tree based interleaver is found to be
considerably less that of random interleaver and slightly higher to that of master
random interleaver. It has also been observed that BER performance of this
interleaver is similar to that of random interleaver and master random interleaver.
In addition to it, the maximal ratio combining (MRC) diversity has also been
incorporated with tree based interleaver along with other interleavers for the purpose
of comparisons. In every case, the BER performance of proposed tree based
interleaver is observed to be very near to that for random interleaver. Further, the
correlation analysis along with FPGA implementation of all the concerned
interleavers have been carried out in this thesis. After the analysis, it has been
119

observed that proposed tree based interleaver requires least hardware along with least
timing constraints for its operation.
Specific contributions made in this thesis are listed below.
A new optimum user-specific interleaver generation mechanism, named as
tree based interleaver (TBI) for IDMA scheme has been suggested. Three different
user-specific interleaver generation mechanisms including RI, MRI, and proposed
TBI have been taken up for analysis and simulation. In this work the discussion and
analysis is concentrated on RI and proposed TBI generation mechanisms. The reason
for opting RI for comparative analysis is inherent in property of its randomness and
hence providing better orthogonality amongst all user-specific interleavers.
Performance evaluation under uniform environment shows that the proposed TBI
generation mechanism has optimum performance to RI and MRI mechanisms, in
terms of memory, computational complexity, and bit error rate (BER) parameters. It is
also observed that bandwidth requirement of proposed TBI mechanism is also
optimum because only two master interleaver and user count has to be transmitted
along with the combined user data. The data formats for all the three interleaving
mechanisms have been duly presented in figure 3.6 in section 3.4. The performance
evaluation under unequal power allocation algorithm again confirms the similar BER
performance of TBI mechanism in comparison to that of RI mechanism.
The proposed tree based interleaver generation mechanism has been incorporated
with maximal ratio combining (MRC) diversity technique. For the purpose of
performance evaluation, both receive and transmit diversity architecture has been
adopted during simulation in uncoded and rate convolutionally coded environment.
On observation of simulation results, it is noticed that the performance in terms of
120

BER for proposed interleaver generation mechanism is found to be very close to that
of RI generation mechanism with variation in respective user count.
The correlation analysis and field-programmable gate-array (FPGA)
implementation of RI, MRI, and proposed TBI mechanisms have also been studied. It
is observed that the cross-correlation of TBI mechanism is almost similar to that of
MRI and RI. It reconfirms that BER performance TBI mechanism should be similar to
that of other schemes. During the FPGA implementations, it is observed that the
hardware requirement of tree based interleaver (TBI) generation mechanism is at the
minimum level in comparison to other considered mechanisms. The timing
constraints of TBI mechanism are also found to be at its minimum level. As the
requirement of looping operation in TBI generation mechanism, is quite less in
comparison to that of MRI, it inherits comparably lesser complexity.
The proposed tree based interleaver (TBI) generation mechanism has shown
optimum overall BER and other performances in comparison to RI, and MRI
mechanisms. The computational complexity and timing constraints for TBI
mechanism are also found to be at minimum level. Hence, the proposed TBI
mechanism is ideally suited for IDMA scheme as its integral part.

6.1 Suggestions for Further Investigations
One of the promising extensions of this study can be the hardware implementation of
IDMA transmitter and receiver system duly implemented with proposed tree based
interleaving mechanism. The parameters can further be compared with all the relevant
parameters of CDMA scheme. It will make the vision more clear regarding hardware
requirements and timing constraints of IDMA systems in comparison to its
counterparts.
121

Another potential extension of the work may be implementation of optically
wired and wireless channels along with appropriate coding techniques. Optical
analysis of IDMA scheme will extend the boundaries of BER performance up to limit
of extreme satisfaction. As optical devices mature and become cost-competitive,
optical IDMA may play a much more significant role wired and wireless
communication systems. Looking at the demand for bandwidth, the vision of using
optical IDMA systems in the communication systems is very promising. This research
will open the new path for future optical communication with IDMA scheme
incorporated with tree based interleavers. The similar analysis can also be carried out
with OFDM-IDMA scheme.
Apart from this, the tree based interleavers can also be implemented with
newly evolved diversity mechanism such as cooperative diversity with IDMA
scheme. In alternate words, it can be stated the sky is limit for research work related
to IDMA systems with tree based interleavers.











122

References


1. http://ezinearticles.com/?Cellular-Market-in-India&id=1599054
2. http://www.india-cellular.com/News.htm
3. http://in.reuters.com/article/domesticNews/idINSGE63Q05A20100427
4. http://apl.jhu.edu/~paulmac/c++-memoization.html
5. International Telecommunication Union, Framework and overall objectives of
the future development of IMT-2000 and systems beyond IMT-2000, Rec. ITU-
R M.1645, June 2003.
6. Pingzhi Fan, Multiple Access Technologies for Next Generation Mobile
Communications, in Proc. 6th International Conference on ITS
Telecommunications, pp. 10-11. 2006.
7. J. G. Andrews, Interference cancellation for cellular systems: A contemporary
overview, IEEE Trans. Wireless Commun., vol. 12, pp. 1929, Apr. 2005.
8. T. S. Rappaport, Wireless Communications: Principles and Practice, 2nd.
Englewood Cliffs, NJ: Prentice-Hall, 2002.
9. D. Tse and P. Viswanath, Fundamentals of Wireless Communication,
Cambridge: Cambridge University Press, 2005.
10. Liberti J.C., and Rappaport T.S., Smart Antennas for Wireless
Communications: IS-95 and Third Generation Applications, Prentice Hall,
Upper Saddle River, NJ, 1999.
11. K. Fazel and S. Kaiser, Multi-Carrier and Spread Spectrum Systems,
Chichester, John Wiley & Sons, 2003.
12. Andrea Goldsmith, Wireless Communications, Cambridge University
Press, 2005.
123

13. X. Ma and L. Ping, Coded modulation using superimposed binary codes, IEEE
Trans. Inform. Theory, vol. 50, pp. 33313343, Dec. 2004.
14. G. N. Karystinos, D. A. Pados, New Bounds on the Total Squared Correlation
and Optimum Design of DS-CDMA Binary Signature Sets IEEE Transactions
on Communications, vol. 51, pp. 48-31, Jan 2003.
15. L.D. Olavarrieta, A.A. Nava, Wireless Communications: A Birds Eye View of
an Emerging Technology, in Proc. IEEE International Symposium on
Communications and Information Technology, ISCIT 2004. , vol.1, pp. 541-546,
2004.
16. P. Wang, J. Xiao, and Li Ping, Comparison of Orthogonal and Non-Orthogonal
Approaches to Future Wireless Cellular Systems, IEEE Vehic. Tech. Mag., vol.
1, pp. 411, Sept. 2006.
17. W.C.Y. Lee Overview of Cellular CDMA, IEEE Transactions on Vehicular
Technology, vol. 40, pp 291-302, May 1991.
18. L.B. Milstein, D.L Schilling, The CDMA Overlay Concept, in Proc. IEEE 4th
International Symposium on Spread Spectrum Techniques and Applications,
vol.2, pp. 476-480, 1996.
19. Gong-Liang Liu, Xiu-Hong Wang, Hui-Xiao Ma, Xing-Peng Mao, IDMA-
Based MAC Protocol for Next-Generation Satellite Networks, in Proc. 5th
International Conference on Wireless Communications, Networking and Mobile
Computing, 2009. WiCom '09, pp. 1-4, 2009.
20. H. Yang, A road to future broadband wireless access: MIMO-OFDM based
interface, IEEE Commun. Mag., vol. 43, pp. 53-60, Jan 2005.
21. K. Yang, X. Wang, A multicarrier chip-interleaved multiuser UWB system, in
Proc. ICASSP05, pp. 325-328, 2005.
124

22. Kai Li, Xiaodong Wang, Guosen Yue and Li Ping, A Low-Rate Code-Spread
and Chip-Interleaved Time-Hopping UWB System, IEEE J. Select. Areas
Commun., vol. 24, pp. 864-870, April 2006.
23. D. Gesbert, M. Shafi, D. Shiu, P.J. Smith, and A. Naguib, From theory to
practice: An overview of MIMO space-time coded wireless systems, IEEE J.
Selected Areas in Commun., vol. 21, pp. 281-302, April 2003.
24. H.Schoeneich, P. Hoeher, Adaptive Interleave-Division Multiple Access- A
Potential Air Interface for 4G Bearer Services and Wireless LANs, in Proc.
WOCN04, Musket , Oman, June 2004.
25. Li Ping, L. Liu, K. Y. Wu, and W. K. Leung, Approaching the capacity of
multiple access channels using interleaved low-rate codes, IEEE Comm. Letter,
vol. 8, pp. 46, Jan. 2004
26. H. Sari, F. Vanhaverbeke, and M. Moeneclaey, Extending the capacity of
multiple access channels. IEEE Comm. Magazine, vol. 38, pp. 74-82, Jan. 2000.
27. F. Vanhaverbeke, M. Moeneclaey, and H. Sari, Increasing CDMA capacity
using multiple sets of orthogonal spreading sequences and soft interference
cancellation, in Proc. IEEE ICC '02, New York, pp. 1516-1520, Apr./May
2002.
28. P. Hoeher, U. Sorger, and I. Land, Log-likelihood values and Monte Carlo
simulation. Some fundamental results, in Proc. 2nd Int. Symp. on Turbo Codes
& Related Topics, Brest, France, pp. 43-46, Sept. 2000.
29. C. Berrou and A. Glavieux, Near Shannon limit error correcting coding and
decoding: Turbo-codes, IEEE Trans. Commun., vol. 44, pp. 12611271, Oct.
1996.
125

30. J. Viterbi, Very low rate convolutional codes for maximum theoretical
performance of spread spectrum multiple-access channels, IEEE J. Select.
Areas Commun., vol. 8, pp. 641649, Aug. 1990.
31. K. Y. Wu, W. K. Leung, and Li Ping, A simple approach to near-optimal
multiple transmit antenna space-time codes, in Proc. IEEE ICC03, Alaska,
USA, pp. 2603-2607, May 2003.
32. K. Y. Wu and Li Ping, Multi-layer turbo space-time codes, IEEE Commun.
Letter, vol. 9, pp. 5557, Jan. 2005.
33. W. K. Leung, K. Y. Wu, and Li Ping, Interleave Division multiplexing Space-
Time Codes, in Proc. ISIT03, Yokohama , Japan, pp.354, June 2003.
34. 34]Stephan Ten Brink, Coding over Space and Time for Wireless Systems,
IEEE wireless communications, pp. 18-30, vol. 13, Aug. 2006.
35. Keying Wu, Li Ping and Jinhong Yuan, A Quasi-Random Approach to Space
Time Codes, in Proc. International Conference on Turbo Coding06, Munich,
April 2006.
36. Vahid Tarokh, Hamid Jafarkhani and A. R. Calderbank, Space-Time Block
Codes from Orthogonal Designs, IEEE Transactions on Information Theory,
vol. 45, pp. 1456-1467, July 1999.
37. Vahid Tarokh, Hamid Jafarkhani and A. R. Calderbank, Space-Time Block
Coding for Wireless Communications: Performance Results, IEEE J. Select.
Areas in Commun., vol. 17, pp. 451-460, Mar. 1999.
38. Vahid Tarokh, Nambi Seshadri and A. R. Calderbank, Space-Time Codes for
High Data Rate wireless Communication: Performance Criterion and Code
Construction, IEEE Transactions on Information Theory, vol. 44, pp. 744-765,
March 1998.
126

39. Li Ping , K. Y. Wu, W. K. Leung, and, A simple unified approach to nearly-
optimal detection and space-time coding, in Proc. ITW02, Bangalore , India,
pp. 53-56, Oct. 2002.
40. Jun Tong, Li Ping and Xiao Ma, Superposition Coding with Peak Power
Limitation, in Proc. IEEE Int. Conf. on Commun., ICC06, Istanbul, Turkey,
Jun. 2006.
41. X. Zhang, Z. Bie, Z. He and W. Wu, Boosts to LDPC-coded IDMA systems,
IEEE Electronics Letters, vol. 44, 2008.
42. Tao Peng; Xiao Yue; Xu He; Shaoqian Li; Simplified receiver for LDPC-Coded
interleave-division multiple-access (IDMA) system, in Proc. International
Conference on Communications, Circuits and Systems, ICCCAS 2009. pp. 108 -
111, 2009.
43. Yao Luo, Lin Wang, Jun Ling, The Coding-Spreading Trade-off in LDPC-
Coded Interleave-Division Multiple-Access (IDMA) System, in Proc.
International Conference on Communication Technology, ICCT '06, pp. 1-6,
2006.
44. 44]Gonzalez-Lopez, Vazquez-Araujo, Castedo, Garcia-Frias, Interleave-
division multiple access (IDMA) using low-rate layered LDGM codes , in Proc.
of 5th International Symposium on Turbo Codes and Related Topics , pp. 315
320, 2008.
45. M. C. Reed, C. B. Schlegel, P. D. Alexander, and J. A. Asenstorfer, Iterative
multi-user detection for CDMA with FEC: Near-single-user performance, IEEE
Trans .on Commun., vol. 46, pp. 16931699, Dec. 1998.
127

46. Abdelkrim, H., Ali, K., Salim, C., Gerard, S., Independent component analysis
in IDMA systems. in Proc. of Joint IEEE North-East Workshop on Circuits
and Systems and TAISA Conference, 2009. NEWCAS-TAISA '09. , pp. 1-4, 2009.
47. Xingzhong Xiong; Jianhao Hu; Lan Tian; A fast converging multi-user
detection for IDMA based on time-reversal , in Proc. 6th International
Conference on Information, Communications & Signal Processing, 2007 , pp. 1-
7, 2007.
48. B. Cristea, D. Roviras, B. Escrig, Turbo receivers for interleave-division
multiple-access systems, IEEE Transactions on Communications, vol. 57, pp.
2090-2097, 2009.
49. S. Houcke, G. Sicot, M. Debbah, Blind Detection for Block Coded Interleaved-
Division Multiple-Access, in Proc. IEEE Global Telecommunications
Conference, GLOBECOM '06 , pp 1-5, 2006.
50. Z. Shi and C. Schlegel, Joint iterative decoding of serially concatenated error
control coded CDMA, IEEE J. Select. Areas Commun., vol. 19, pp. 16461653,
Aug. 2001.
51. AlRustamani, A. D. Damnjanovic, and B. R. Vojcic, Turbo greedy multi-user
detection, IEEE J. Select. Areas Commun., vol. 19, pp. 16381645, Aug. 2001.
52. Lihai Lui and Li Ping, A Comparative Study on Low-Cost Multiuser
Detectors, in Proc. IEEE Int. Conf. on Commun. ICC06, Istanbul, Turkey,
2006.
53. R. J. McEliece, Are turbo-like codes effective on nonstandard channels? IEEE
Inform. Theory Society Newsletter, vol. 51, pp. 18, Dec.2001.
128

54. M. C. Reed and P. D. Alexander, Iterative multi-user detection using antenna
arrays and FEC on multipath channels, IEEE J. Select. Areas Commun., vol. 17,
pp. 20822089, Dec. 1999.
55. S. Verd and S. Shamai, Spectral efficiency of CDMA with random spreading,
IEEE Trans. Inform. Theory, vol. 45, pp. 622640, Mar. 1999.
56. J. Mietzner and P.A. Hoeher, Boosting the performance of wireless
communication systems: theory and practice of multiple-antenna techniques,
IEEE Commun. Mag., vol. 42, pp. 40-47, Oct 2004.
57. M. Moher and P. Guinand, An iterative algorithm for asynchronous coded
multi-user detection, IEEE Commun. Lett., vol. 2, pp. 229231, Aug. 1998.
58. Javed, S., Khokhar, I.A., Obaid, A., A novel proposal for chip-by-chip iterative
detection in Interleave-division multiple access (IDMA) in Proc. International
Symposium on High Capacity Optical Networks and Enabling Technologies,
HONET 2007 , pp. 1-3, 2007.
59. O. Nagy, M.C. Reed, Zhenning Shi; Optimal Detection of IDMA Signals , in
Proc. of IEEE Conf. Wireless Communications and Networking Conference,
WCNC 2007, pp. 1236 1240, 2007.
60. Rong Zhang, Lei Xu, Sheng Chen, L. Hanzo, EXIT-Chart-Aided Hybrid
Multiuser Detector for Multicarrier Interleave-Division Multiple Access, IEEE
Transactions on Vehicular Technology, vol. 59, pp. 1563-1567, 2010.
61. F. N. Brannstrom, T. M. Aulin, and L. K. Rasmussen, Iterative decoders for
trellis code multiple-access, IEEE Trans. on Communs., vol. 50, pp. 14781485,
Sept. 2002.
129

62. X. Wang and H. V. Poor, Iterative (turbo) soft interference cancellation and
decoding for coded CDMA, IEEE Trans. Commun., vol. 47, pp. 10461061,
July 1999.
63. L. Liu, W. K. Leung, and Li Ping, Simple chip-by-chip multi-user detection for
CDMA systems, in Proc. IEEE Conference on Vehicular Technology, VTC-
Spring, Korea, pp. 2157-2161, Apr. 2003.
64. G. Caire, S. Guemghar, A. Roumy, and S. Verdu, Maximizing the spectral
efficiency of coded CDMA under successive decoding, IEEE Inform. Theory,
vol. 50, pp.152164, Jan. 2004.
65. Li Ping, L. Liu, Iterative detection of chip interleaved CDMA systems in
multipath channels, in IEEE Electronics Letters, vol.40, July 2004.
66. Shu-Ming Tseng, Yu-Chin Kuo, Deterministic IDMA and Deterministic
CDMA: A comparison, in Proc. 14th Asia-Pacific Conference on
Communications, APCC 2008. , pp. 1-3, 2008.
67. Basharat A., Khokhar, I.A., Murtaza, S., CDMA versus IDMA for subscriber
cell density, in Proc. International Conference on Innovations in Information
Technology, IIT 2008. , pp. 520-524, 2008.
68. Katsutoshi Kusume, Gerhard Bauch, Wolfgang Utschick, IDMA vs. CDMA:
Detectors, Performance and Complexity, in Proc. IEEE Global
Telecommunications Conference, 2009. GLOBECOM 2009, pp. 1-8, 2009.
69. Yifeng Tu, Pingzhi Fan, Gao Zhou, Grouped Interleave-Division Multiple
Access, in Proc. First International Conference on Communications and
Networking in China, ChinaCom '06, pp. 1-5, 2006.
130

70. Senanayake09]B. Senanayake, M.C. Reed, Zhenning Shi,
An Optimal Asynchronous IDMA Receiver, in Proc. Australian
Communications Theory Workshop, 2009. AusCTW 2009, pp. 28-32, 2009.
71. Li Ping, L. Liu, K. Y. Wu, and W. K. Leung, A unified approach to multi-user
detection and space time coding with low complexity and nearly optimal
performance, in Proc. 40th Allerton Conference, Allerton House, USA, pp.
170179, Oct. 2002.
72. Li Ping, Lihai Lui and W .K. Leung, A Simple Approach to Near-Optimal
Multiuser Detection: Interleave-Division Multiple-Access, IEEE, 2003.
73. Yi Hong, L.K Rasmussen, Iterative Switched Decoding for Interleave-Division
Multiple-Access Systems, IEEE Transactions on Vehicular Technology, vol.
57, pp. 1939-1944, 2008.
74. H. Samir, K. Noori, S.A. Haider, S. Sultan, Implementation design and analysis
of coded and un-coded multi- user IDMA system , in Proc 6th International
Symposium on High-Capacity Optical Networks and Enabling Technologies
(HONET), pp. 125-128, 2009.
75. Rong Zhang, L. Hanzo, Three Design Aspects of Multicarrier Interleave
Division Multiple Access, IEEE Transactions on Vehicular Technology, vol.
57, pp. 3607-3617, 2008.
76. Li Ping, Lihai Liu, Keying Wu, W. Leung, Interleave Division Multiple
Access, IEEE Transactions on Wireless Communications, vol. 5, pp. 938-947,
April 2006
77. L. Ping, L. Liu, K. Wu, and W. Leung, On interleave-division multiple access,
in Proc. ICC 2004, pp. 28692873, June 2004.
131

78. Xingzhong Xiong, Jianhao Hu, Xiang Ling, IDMA/CDMA: A Hybrid Multiple
Access Scheme, in Proc. of International Conference on MultiMedia and
Information Technology, MMIT '08. , pp. 512-514, 2008.
79. Li Ping, Wang Peng, Wang Xiaodong, Recent Progress in Interleave-Division
Multiple-Access (IDMA) , in Proc. IEEE Military Communications
Conference, 2007. MILCOM 2007, pp. 1-7, 2007.
80. Li Ping and Lihai Liu, Analysis and design for IDMA systems based on SNR
evolution and power allocation, in Proc. IEEE VTC04 Fall, Los Angels, CA,
vol. 2, pp.1068-1072, Sept. 2004.
81. Zvi Rosberg, Optimal Transmitter Power Control in Interleave Division
Multiple Access (IDMA) Spread Spectrum Uplink Channels, IEEE
Transactions on Wireless Communications, vol. 5, 2007.
82. M.S.K. Lau, Wuyi Yue, Peng Wang, Li Ping, A Branch-and-Bound Method for
Power Minimization of IDMA, IEEE Transactions on Vehicular Technology,
vol. 57, pp. 3525 3537, 2008.
83. Peng Wang, Li Ping and Lihai Liu, Power Allocation for Multiple Access
Systems with Practical Coding and Iterative Multi-User Detection, in Proc.
Research Grant Council, Hongkong SAR, China.,2006.
84. P. Weitkemper, K.D. Kammeyer, Power Optimization of IDMA Systems with
Different Target BER Constraints, in Proc. IEEE 65th Vehicular Technology
Conference, VTC2007-Spring. , pp. 2812 2816, 2007.
85. Kai Li, Xiaodong Wang and Li Ping, Analysis and Optimization of Interleave-
Division Multiple Access Communication System, in Proc. ICASS, vol. 3, pp.
917-920, 2005.
132

86. Kai Li, Xiaodong Wang and Li Ping, Analysis and Optimization of Interleave-
Division Multiple Access Communication System, IEEE Transactions on
Wireless Communications, vol. 6, 2007.
87. Peng Wang, Li Ping and Lihai Liu, Optimized Power Allocation for Multiple
Access Systems with Practical Coding and Iterative Multiuser Detection, in
Proc. Int. Symp. Turbo codes & Related Topics, Munich, Germany, Aril 3-7,
2006.
88. Peng Wang, Li Ping and Lihai Liu, Power Allocation for Multiple Access
Systems with practical coding and Iterative Multiuser Detection, in Proc. IEEE
Int. Conf. on Commun., ICC06, Istanbul, Turkey, June 11-15, 2006.
89. L. Linton, P. Conder, M. Faulkner, Multi-Rate Communications Using Layered
Interleave-Division Multiple Access with Power Allocation, in Proc. IEEE
Conference on Wireless Communications and Networking, WCNC 2009 , pp. 1-
5, 2009.
90. M.S.K. Lau, Wuyi Yue, Li Ping, Equal power allocation of IDMA systems:
Feasibility, optimality, and throughput IEEE Transactions on Wireless
Communications, vol. 8 , pp. 746 756, 2009.
91. I. Pupeza, A. Kavcic, and L. Ping, Efficient generation of interleavers for
IDMA, in Proc. IEEE International Conference on Communications, ICC 2006,
vol. 4, pp. 1508-1513, June 2006.
92. C. Zhang and J. Hu, The shifting interleaver design based on PN sequence for
IDMA systems, in Proc. of International Conference on Future Generation
Communication and Networking, FGCN 2007, Korea, December 2007.
133

93. C. Zhang and J. Hu, 2-dimension interleaver design for IDMA systems, in
Proc. IEEE International Conference on Circuits and Systems for
Communications, ICCSC 2008, pp. 372-376, May 2008.
94. K. Kusume, G. Bauch, Simple construction of multiple interleavers: cyclically
shifting a single interleaver , IEEE Transactions on Communications, vol. 56,
pp. 1394-1397, 2008.
95. Mark S. C. Ho, Steven S. Pietrobon and Tim Giles, Interleavers for Punctured
Turbo Codes, Telstra Australia under Contract No. 700136, Nov. 1998.
96. J. Hokfelt, O. Edfors and T. Maseng, A Turbo Code Interleaver Design
Criterion Based on the Performance of Iterative Decoding, IEEE
Communications Letters, vol. 5, Feb. 2001.
97. Lihai Liu, Jun Tong and Li Ping, Analysis and Optimization of CDMA systems
with chip level Interleavers, IEEE J. Select. Areas Commun., vol. 24, pp. 141-
150, Jan. 2006.
98. Tarable, G. Montorsi, and S. Benedetto, Analysis and design of interleavers for
CDMA systems, IEEE Commun. Lett., vol. 5, pp. 420422, Oct. 2001.
99. S. Brck, U. Sorger, S. Gligorevic, and N. Stolte, Interleaving for outer
convolutional codes in DSCDMA Systems, IEEE Trans. Commun., vol. 48, pp.
11001107, July 2000.
100. Shuang Wu, Xiang Chen, Shidong Zhou, A Parallel Interleaver Design for
IDMA Systems, in Proc. International Conference on Wireless
Communications & Signal Processing, WCSP 2009, pp.1-5. 2009.
101. Li Han, Minglu Jin, Ende Song, Matrix Cyclic Shifting Based Interleaver
Design for IDMA System , in Proc. 5th International Conference on Wireless
134

, pp. 1-4,
2009.
102. Yeon Ho Chung, Performance and bandwidth efficient interleave-division
multiple access scheme with high-spread interleavers, in Proc. 6th
International Conference on Information, Communications & Signal
Processing, pp. 1-5, 2007.
103. R. H. Mahadevappa and J. G. Proakis, Mitigating multiple access
interference and intersymbol Interference in uncoded CDMA Systems with chip
level interleaving, IEEE Trans. Wireless Commun., vol. 1, pp. 781792, Oct.
2002.
104. H. Wu, L.Ping and A. Perotti, User-specific chip-level interleaver design for
IDMA System, IEEE Electronics Letters, vol.42, Feb 2006.
105. Xinyi Xu; Qiuping Zhu, Liang Zhu, Weibing Wang; Kegang Liu, The Model
of Evolutionary Interleavers for IDMA Communication System in Proc.
International Conference on Wireless Communications, Networking and Mobile
Computing, WiCom 07, pp. 751 754, 2007.
106. Zliisong Bie, Weiling Wu, PEG Algorithm Based Interleavers Design for
IDMA System in Proc. 41st Annual Conference on Information Sciences and
Systems, CISS '07 , pp. 480 483, 2007.
107. D. Hao, P. Hoeher, Helical interleaver set design for interleave-division
multiplexing and related techniques IEEE Communications Letters , vol. 12,
pp. 843-845, 2008.
108. K. Kusume, G. Bauch, Cyclically Shifted Multiple Interleavers, in Proc.
IEEE Global Telecommunications Conference, GLOBECOM '06 , pp. 1-5, 2006.
135

109. Dapeng Hao, Pin Yao, P.A. Hoeher, Analysis and design of interleaver sets
for interleave-division multiplexing and related techniques, in Proc. 5th
International Symposium on Turbo Codes and Related Topics, pp. 432-437,
2008.
110. A. Mukherjee, H.M. Kwon, Multicarrier Interleave-Division Multiple-Access
Systems with Adaptive Pilot-Based User Interleavers, in Proc. IEEE 70th
Vehicular Technology Conference Fall (VTC 2009-Fall), pp. 1-5, 2009.
111. Zhifeng Luo, Wong, A.K., Shuisheng Qiu, Interleaver design based on linear
congruences for IDMA systems, in Proc. International Conference on IEEE
10th Annual Wireless and Microwave Technology Conference, 2009.
WAMICON '09, pp. 1-4, 2009.
112. Xiaoming Peng, AS Madhukumar , Francois P.S. Chin Performance Studies
of Interleaving Schemes for MC-CDMA Systems, in Proc. PWCNC 2004,
IEEE Communications Society, pp. 2081-2086, 2004.
113. S.M. Alamoutis "A Simple transmitter diversity scheme for wireless
comm., IEEE J. Select Areas Commun., vol 16, pp 1451-1458, Oct 1998.
114. L. Linton, P. Conder, M. Faulkner, Multiuser MIMO Communications using
Interleave-Division Multiple-Access and Golden Codes, in Proc. IEEE
Vehicular Technology Conference, VTC Spring 2008 , pp. 842-846, 2008.
115. Bin Zhang, Kai Niu, Zhiqiang He, Relaying Scheme Based on IDMA for
MIMO Relay Networks , in Proc. 5th International Conference on Wireless
Communications, Networking and Mobile Computing, WiCom '09, pp. 1-4, 2009.
116. S. W. Lei, Vincent K. N. Lau, Performance Analysis of Adaptive
Interleaving for OFDM systems, IEEE Trans. on Vehicular Technology, vol. 51,
May 2002.
136

117. N. Yee, J. P. Linnartz and G. Fettweis, Multi-Carrier CDMA in Indoor
Wireless Radio Networks, in Proc. Proceedings of PIMRC 93, Yokohama,
Japan, pp. 109-113, 1993.
118. H. Atarashi, M. Sawahashi, Variable Spreading Factor Orthogonal Frequency
and Code Division Multiplexing (VSFOFCDM), , in Proc. Third International
workshop on MC-SS 2001 & Related Topics, pp. 203-207, Sept. 2001.
119. Jun Tong, Qinghua Guo and Li Ping Analysis and design of OFDM-IDMA
systems, Eur. Trans. on Telecomms. 2008; pp. 561569, June, 2008.
120. K. Kusume and G. Bauch, CDMA and IDMA: Iterative Multiuser Detections
for Near-Far Asynchronous Communications, IEEE Int. Symp. Personal,
Indoor, and Mobile Radio Commun. (PIMRC05), Berlin, Germany, vol. 1, pp.
42631, Sept. 2005.
121. Suyama, S., Li Zhang, Suzuki, H., Fukawa, K., Performance of Iterative
Multiuser Detection with Channel Estimation for MC-IDMA and Comparison
with Chip-Interleaved MC-CDMA, in Proc. IEEE Global Telecommunications
Conference, 2008. IEEE GLOBECOM 2008, pp. 1-4, 2008.
122. Shanshan Yang, Weikai Xu, Lin Wang, Qinfang Wei, Performance of STBC-
IDMA system over quasi-static Rayleigh fading channel, in Proc. International
Conference on Communications, Circuits and Systems, ICCCAS 2008, pp. 44-
77, 2008.
123. Novak, C., Hlawatsch, F., Matz, G., MIMO-IDMA: Uplink Multiuser MIMO
Communications using Interleave-Division Multiple Access and Low-
Complexity Iterative Receivers, in Proc. IEEE International Conference on
Acoustics, Speech and Signal Processing, ICASSP 2007, pp. III-225 - III-228,
2007.
137

124. Chulhee Jang, Jae Hong Lee, IDMA system with relays, in Proc. 1st
International Conference on Wireless Communication, Vehicular Technology,
Information Theory and Aerospace & Electronic Systems Technology, Wireless
VITAE 2009. , pp. 404 408, 2009.
125. P. Weitkemper, D. Wubben, K. Kammeyer, Delay-diversity in multi-user
relay systems with Interleave Division Multiple Access,
in Proc. IEEE International Symposium on Wireless Communication Systems,
ISWCS '08 , pp. 364-368, 2008.
126. Zhifeng Luo, Gurkan, D., Zhu Han, Wong, A.K., Shuisheng Qiu,
Cooperative Communication Based on IDMA , in Proc. 5th International
Conference on Wireless Communications, Networking and Mobile Computing,
WiCom '09, pp. 1-4, 2009.
127. Xingzhong Xiong, Jianhao Hu, A simplified downlink transmission and
receiving scheme for IDMA, in Proc. International Conference on
Communications, Circuits and Systems, ICCCAS 2008, pp. 157 -161, 2008.
128. Xiangyun Zhou; Zhenning Shi; Reed, M.C.; Iterative Channel Estimation for
IDMA Systems in Time-Varying Channels, in Proc. IEEE Conference on
Global Telecommunications Conference, 2007. GLOBECOM 07, pp. 4020-
4024, 2007.
129. A. Basharat, I. A. Khokhar, S. Murtaza, Security Analysis of Interleave-
Division Multiple Access: Next Generation Competitor, in Proc.
The Fourth International Conference on Wireless and Mobile Communications,
ICWMC '08, pp. 274-278, 2008.
130. Xingzhong Xiong; Jianhao Hu; Performance Analysis of Anti-jamming for
Interleave-Division Multiple Access Systems, in Proc. International
138

Wireless Networks and Information Systems, WNIS '09., pp. 15-
18, 2009.
131. Xingzhong Xiong; Jianhao Hu; Feng Yang; Xiang Ling; Effect of channel
estimation error on the performance of Interleave-Division Multiple Access
systems , in Proc. 11th International Conference on Advanced Communication
Technology, ICACT 2009. , pp. 1538-1542, 2008.
132. Ur Rehman, Imran Zaka, M. Naeem, S.I. Shah, J. Ahmad, Multicode
Multicarrier Interleave Division Multiple Access Communication, in Proc.
IEEE Multitopic Conference, INMIC '06 , pp. 37-41, 2006.
133. Dageng Chen; Yi Wang; Jiayin Zhang; An Interleave-Division multiplexing
(IDM) based modulation for 3GPP LTE downlink , in Proc. Third International
Conference on Communications and Networking in China, ChinaCom 2008 , pp.
474-477, 2008.
134. Qian Huang; King-Tim Ko; Peng Wang; Li Ping; Chan, S.;
Interleave-Division Multiple-Access Based Broadband Wireless Networks , in
Proc. IEEE Information Theory Workshop, 2006. ITW '06, pp. 502-506, 2006.
135. Hoeher, P.A.; Wen Xu; Multi-Layer Interleave-Division Multiple Access for
3GPP Long Term Evolution, in Proc. IEEE International Conference on
Communications, ICC '07, pp. 5508-5513, 2007.
136. Feng Yang, Jianhao Hu, Xingzhong Xiong, EM Combination Channel
Estimation Algorithm for Interleave-Division Multiple Access Systems, in
Proc. International Conference on MultiMedia and Information Technology,
2008. MMIT 08, pp. 515-518, 2008.
139

137. Jong-Bu Lim, Chan-Ho Choi, Tae-Won Yune, Gi-Hong Im, Iterative
Multiuser Detection for Single-Carrier Modulation with Frequency-Domain
Equalization, IEEE Communications Letters, vol. 11, pp. 471-473, 2007.
138. Jianzhong Huang; Yang Yang; Xinmei Wang; Peng Wang; Li Ping;
Transmission Strategies for Parallel Relay Networks Based on Superposition
Coding , in Proc. IEEE 68th Vehicular Technology Conference, 2008. VTC
2008-Fall, pp.1-5, 2008.
139. Bernard Sklar Rayleigh Fading Channels in Mobile Digital Communication
Systems Part II: Mitigation IEEE Communication magazine, pp 148-155, Sept.
1997.
140. K.S. Andrews, D. Divsalar, S. Dolinar, J. Hamkins, C.R. Jones, F. Pollara,
The Development of Turbo and LDPC Codes for Deep-Space Applications ,
Proceedings of the IEEE , vol. 95, pp. 2142-2156, 2007.
141. Nirmal Chandrasekaran, Diversity Techniques in Wireless Communication
Systems Wireless Communication Technologies Spring, 2005.
142. B. Hongxia , B. Zhisong, A hybrid multiple access scheme: OFDMA-IDMA,
in Proc. First International Conference on Communications and Networking,
China, ChinaCom '06, pp. 1-3, 2006.
143. Li Ping, Qinghua Guo, Jun Tong, The OFDM-IDMA approach to wireless
communication systems, IEEE Wireless Communication, vol. 14, issue 4, pp.
18-24, 2007.
144. C. Novak, G. Matz, F. Hlawatsch, Factor graph based design of an OFDM-
IDMA receiver performing joint data detection, channel estimation, and channel
length selection, In Proc. IEEE International Conference on Acoustics,
Speech and Signal Processing, 2009. ICASSP 2009. , pp. 2561-2564, 2007.
140

145. B. K. Upadhyaya, S. K. Sanyal, VHDL Modeling of Convolutional
Interleaver-Deinterleaver for Efficient FPGA Implementation In Proc.
International Journal of Recent Trends in Engineering, Vol 2, No. 6, pp. 66-28,
2009.
146. Gianluca Gilardi, Catello Antonio De Rosa, Designing Convolutional
Interleavers with Virtex Devices, Application notes of Vertex series, Xilinx, pp.
1-6, 2000.
147. M. Shukla, V.K. Srivastava, S. Tiwari, Analysis and Design of Optimum
Interleaver for Iterative Receivers in IDMA Scheme, Wiley Journal of Wireless
Communication and Mobile Computing, Vol 9. Issue 10, pp. 1312-1317. Oct,
2009.
148. http://www.imaworld.org/uploads/dbsAttachedFiles/01.Pdf













141

Appendix

List of Publications



Journals


1. M. Shukla, V.K. Srivastava, S. Tiwari, Analysis and Design of Optimum
Interleaver for Iterative Receivers in IDMA Scheme, Wiley Journal of
Wireless Communications and Mobile Computing Vol 9. Issue 10, pp.
1312-1317, 2009.
2. M. Shukla, Aasheesh Shukla, Rohit Kumar, V.K. Srivastava, Sudarshan
Tiwari, Simple Diversity Scheme for IDMA Communication System,
International Journal of Applied Engineering Research, Vol. 4 No. 6, pp. 877-
883, 2009.
3. M Shukla, V K Srivastava and S Tiwari, A VHDL Implementation of
Orthogonal Interleavers for Interleave Division Multiple Access (IDMA)
Scheme, The IUP Journal of Telecommunications, Vol. I, No. 2, pp. 63-71,
2009.

International Conferences

4. M. Shukla, V.K. Srivastava, S. Tiwari Interleave Division Multiple Access
for Wireless Communication, 'International Conference on Next Generation
Communication Systems: A Perspective, ICONGENCOM 06, J.K.
Institute, Allahabad, India, pp. 150-154, Dec 9-11, 2006.
142

5. M. Shukla, V.K. Srivastava, S. Tiwari Analysis and Design of Tree Based
Interleaver for Multiuser Receivers in IDMA Scheme, 16
th
IEEE
International Conference on Networks ICON 2008, Delhi, India, pp. 1-4,
Dec. 13-14, 2008.
6. M. Shukla, R.C.S. Chauhan, Ruchir Gupta, V.K. Srivastava, S. Tiwari
Performance Analysis of Tree Based Interleaver with Iterative IDMA
Receivers using Optimum Power Allocation Algorithm, First UK-India IEEE
International Workshop on Cognitive Wireless Systems held in IIT, Delhi,
India, Dec. 11-12, 2009.
7. M. Shukla, V.K. Srivastava, S. Tiwari Performance Analysis of Tree Based
Interleaver with IDMA Systems using Optimum Power Allocation
Algorithm, 2nd IEEE International Conference on Emerging Trends in
Engineering & Technology ICETET-09, Nagpur, India, Dec. 16-18, 2009.
8. M. Shukla, Aasheesh Shukla,, V.K. Srivastava, S. Tiwari Performance
Evaluation of MRC Diversity Scheme for Iterative IDMA Receivers , Annual
IEEE India Conference INDICON-09, Gandhinagar, Gujrat, India, pp. 1-4,
Dec. 18-20, 2009.
9. M. Shukla, Aasheesh Shukla,, V.K. Srivastava, S. Tiwari Different
Designing Factors for IDMA Systems, 1
ST
International Conference on
Computer, Communication, and Control and Information Technology C3 IT
2009 in Academy of Technology, Calcutta, India, pp. 748-756, Feb. 6-7,
2009.
10. M. Shukla, P. S. Sharma, V.K. Srivastava, S. Tiwari Performance
Analysis of Iterative IDMA Systems with MRC Diversity in Multipath Fading
Environment using Optimum Tree Based Interleaver, IEEE International
143

144

Conference on Computational Intelligence and Computing Research (ICCIC),
scheduled in Tamilnadu College of Engineering, Coimbatore, India, Dec. 28-
29, 2010.

National Conferences
11. Manoj Shukla, Aasheesh Shukla, V.K. Srivastava, S. Tiwari Interleave
Division Multiple Access Scheme: An Overview, 3
rd
National Conference on
Currents Trends in Technology, Nirma University, Ahmadabad, India, pp.
350-355, Nov. 27-29., 2008.

Communicated Paper
12. M. Shukla, V.K. Srivastava, S. Tiwari FPGA Implementation of Orthogonal
Interleavers for IDMA Scheme, Elsevier International Journal on Computers
& Electrical Engineering (CEE), (Communicated)