Title page for ETD etd-09272009-153908


Document Type Doctoral Thesis
Author Khan, Salman Ahmad
Email salmank@kfupm.edu.sa
URN etd-09272009-153908
Document Title Design and analysis of evolutionary and swarm intelligence techniques for topology design of distributed local area networks
Degree PhD
Department Computer Science
Supervisor
Advisor Name Title
Prof A P Engelbrecht Supervisor
Keywords
  • stochastic evolution
  • swarm intelligence
  • simulated evolution
  • fuzzy logic
  • unified and-or operator
  • optimization
  • local area networks
  • simulated annealing
  • ant colony optimization
  • particle swarm optimizati
Date 2009-09-02
Availability unrestricted
Abstract
Topology design of distributed local area networks (DLANs) can be classified as an NP-hard problem. Intelligent algorithms, such as evolutionary and swarm intelligence techniques, are candidate approaches to address this problem and to produce desirable solutions. DLAN topology design consists of several conflicting objectives such as minimization of cost, minimization of network delay, minimization of the number of hops between two nodes, and maximization of reliability. It is possible to combine these objectives in a single-objective function, provided that the trade-offs among these objectives are adhered to. This thesis proposes a strategy and a new aggregation operator based on fuzzy logic to combine the four objectives in a single-objective function. The thesis also investigates the use of a number of evolutionary algorithms such as stochastic evolution, simulated evolution, and simulated annealing. A number of hybrid variants of the above algorithms are also proposed. Furthermore, the applicability of swarm intelligence techniques such as ant colony optimization and particle swarm optimization to topology design has been investigated. All proposed techniques have been evaluated empirically with respect to their algorithm parameters. Results suggest that simulated annealing produced the best results among all proposed algorithms. In addition, the hybrid variants of simulated annealing, simulated evolution, and stochastic evolution generated better results than their respective basic algorithms. Moreover, a comparison of ant colony optimization and particle swarm optimization shows that the latter generated better results than the former.

2009, 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:

Khan, SA 2009, Design and analysis of evolutionary and swarm intelligence techniques for topology design of distributed local area networks, PhD thesis, University of Pretoria, Pretoria, viewed yymmdd < http://upetd.up.ac.za/thesis/available/etd-09272009-153908/ >

D678/ag

Files
  Filename       Size       Approximate Download Time (Hours:Minutes:Seconds) 
 
 28.8 Modem   56K Modem   ISDN (64 Kb)   ISDN (128 Kb)   Higher-speed Access 
  00front.pdf 139.13 Kb 00:00:38 00:00:19 00:00:17 00:00:08 < 00:00:01
  01chapters1-2.pdf 526.00 Kb 00:02:26 00:01:15 00:01:05 00:00:32 00:00:02
  02chapters3-4.pdf 300.35 Kb 00:01:23 00:00:42 00:00:37 00:00:18 00:00:01
  03chapters5-6.pdf 651.83 Kb 00:03:01 00:01:33 00:01:21 00:00:40 00:00:03
  04chapters7-8.pdf 622.14 Kb 00:02:52 00:01:28 00:01:17 00:00:38 00:00:03
  05chapters9-11.pdf 304.40 Kb 00:01:24 00:00:43 00:00:38 00:00:19 00:00:01
  06back.pdf 230.73 Kb 00:01:04 00:00:32 00:00:28 00:00:14 00:00:01

Browse All Available ETDs by ( Author | Department )

If you have more questions or technical problems, please Contact UPeTD.