Document Type Master's Dissertation Author Myburgh, Hermanus Carel URN etd-07202010-170128 Document Title Low complexity iterative MLSE equalization in extremely long Rayleigh fading channels Degree MEng Department Electrical, Electronic and Computer Engineering Supervisor
Advisor Name Title Prof J C Olivier Committee Chair Keywords
- Rayleigh fading
- low complexity
- Hopfield neural network
Date 2010-03-18 Availability restricted Abstract
In mobile wireless communication systems, the transmitted signal is subjected to various impediments, among which intersymbol interference (ISI) poses a major challenge to communication system designers. ISI is the result of the dispersive nature of a wireless communication channel, causing multiple delayed copies of the original transmitted signal to arrive at the receiver. Since the communication channel acts like a finite impulse response (FIR) filter on the transmitted data, the effect of the channel needs to be reversed in order to reconstruct the information contained in the original transmitted signal. The process of reversing the effect of the channel on the transmitted signal is known as detection or equalization. The computational complexity of conventional optimal equalizers is linear in the length of the transmitted data block and exponential in the channel impulse response (CIR) length. When the bandwidth of the communication signal is low, the duration between the first and the last arrival of the same signal, known as the channel delay spread, only spans a small number of symbols, which implies that the CIR is short. In these systems conventional equalizers can be used to optimally mitigate the effect of ISI. In high bandwidth communication systems, however, the channel delay spread may span tens to hundreds of symbols, which implies very long CIR lengths, rendering conventional equalizers infeasible due to severe computational strain. In this thesis a low complexity maximum likelihood sequence estimation (MLSE) equalizer is developed which has computational complexity quadratic in the length of the transmitted data block and approximately independent of the CIR length for practical single-carrier communication systems. This equalizer is therefore able to equalize signals in systems with hundreds of interfering symbols, at a fraction of the computational complexity of conventional equalizers. Using the Hopfield neural network (HNN) as foundation, this equalizer is applied to underwater communication systems, code division multiple access (CDMA) communication systems, multiple antenna communication systems, and Turbo Equalization, with great success in this dissertation.
Copyrightę 2010, 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.
Please cite as follows:
Myburgh, HC 2010, Low complexity iterative MLSE equalization in extremely long Rayleigh fading channels, MEng dissertation, University of Pretoria, Pretoria, viewed yymmdd < http://upetd.up.ac.za/thesis/available/etd-07202010-170128/ >C10/434/ag
Filename Size Approximate Download Time (Hours:Minutes:Seconds)
28.8 Modem 56K Modem ISDN (64 Kb) ISDN (128 Kb) Higher-speed Access dissertation.pdf 1.40 Mb 00:06:29 00:03:20 00:02:55 00:01:27 00:00:07indicates that a file or directory is accessible from the campus network only.