Digital technology has made complex signal processing possible in communication systems and greatly improved the performance and quality of most modern telecommunication systems. The telecommunication industry and specifically mobile wireless telephone and computer networks have shown phenomenal growth in both the number of subscribers and emerging services, resulting in rapid consumption of common resources of which the electromagnetic spectrum is the most important. Technological advances and research in digital communication are necessary to satisfy the growing demand, to fuel the demand and to exploit all the possibilities and business opportunities. Efficient management and distribution of resources facilitated by state-of-the-art algorithms are indispensable in modern communication networks. The challenge in communication system design is to construct a system that can accurately reproduce the transmitted source message at the receiver. The channel connecting the transmitter and receiver introduces detrimental effects and limits the reliability and speed of information transfer between the source and destination. Typical channel effects encountered in mobile wireless communication systems include path loss between the transmitter and receiver, noise caused by the environment and electronics in the system, and fading caused by multiple paths and movement in the communication channel. In multiple access systems, different users cause interference in each other’s signals and adversely affect the system performance. To ensure reliable communication, methods to overcome channel effects must be devised and implemented in the system. Techniques used to improve system performance and capacity include temporal, frequency, polarisation and spatial diversity. This dissertation is concerned mainly with temporal or time diversity. Channel coding is a temporal diversity scheme and aims to improve the system error performance by adding structured redundancy to the transmitted message. The receiver exploits the redundancy to infer with greater accuracy which message was transmitted, compared with uncoded systems. Sparse graph codes are channel codes represented as sparse probabilistic graphical models which originated in artificial intelligence theory. These channel codes are described as factor graph structures with bit nodes, representing the transmitted codeword bits, and bit-constrained or check nodes. Each constraint involves only a small number of code bits, resulting in a sparse factor graph with far fewer connections between bit and check nodes than the maximum number of possible connections. Sparse graph codes are iteratively decoded using message passing or belief propagation algorithms. Three classes of iteratively decodable channel codes are considered in this study, including low-density parity-check (LDPC), Turbo and repeat-accumulate (RA) codes. The modulation platform presented in this dissertation is a spectrally efficient wideband system employing orthogonal complex spreading sequences (CSSs) to spread information sequences over a wider frequency band in multiple modulation dimensions. Special features of these spreading sequences include their constant envelopes and power output, providing communication range or device battery life advantages. This study shows that multiple layer modulation (MLM) can be used to transmit parallel data streams with improved spectral efficiency compared with single-layer modulation, providing data throughput rates proportional to the number of modulation layers at performances equivalent to single-layer modulation. Alternatively, multiple modulation layers can be used to transmit coded information to achieve improved error performance at throughput rates equivalent to a single layer system / Dissertation (MEng (Electronic Engineering))--University of Pretoria, 2007. / Electrical, Electronic and Computer Engineering / unrestricted
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:up/oai:repository.up.ac.za:2263/26054 |
| Date | 04 July 2007 |
| Creators | Vlok, Jacobus David |
| Contributors | Prof L P Linde, Mr L Staphorst, jdvlok@gmail.com |
| Source Sets | South African National ETD Portal |
| Detected Language | English |
| Type | Dissertation |
| Rights | © 2007, University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria. No part of this work may be reproduced or transmitted in any form or by any means, without the prior written permission of the University of Pretoria. |
Page generated in 0.002 seconds