Please use this identifier to cite or link to this item:
https://hdl.handle.net/2440/17844
Citations | ||
Scopus | Web of Science® | Altmetric |
---|---|---|
?
|
?
|
Type: | Journal article |
Title: | An analytic modelling approach for network routing algorithms that use "ant-like" mobile agents |
Author: | Bean, N. Costa, A. |
Citation: | Computer Networks, 2005; 49(2):243-268 |
Publisher: | Elsevier Science BV |
Issue Date: | 2005 |
ISSN: | 1389-1286 1872-7069 |
Statement of Responsibility: | Nigel Bean and Andre Costa |
Abstract: | In this paper, we introduce an analytic modelling approach to the study of a novel class of adaptive network routing algorithm, which is inspired by the emergent problem-solving behaviours observed in biological ant colonies. This class of algorithm utilizes “ant-like” agents which traverse the network and collectively construct routing policies. Previous studies have focused exclusively on simulation experiments, which indicate that such algorithms perform well in response to real-time changes in traffic demands and network conditions. The analytic model presented in this paper permits useful insights into certain fundamental aspects of ant-based algorithms, which have not been discussed in previous ant-based routing literature. In particular, the work presented in this paper motivates our proposal of a number of modifications to the basic design of ant-based routing algorithms, which result in improved performance with respect to equilibrium performance measures. |
Keywords: | Adaptive routing Ant-based routing Reinforcement learning Wardrop equilibria |
Rights: | © 2005 Elsevier B.V. All rights reserved. |
DOI: | 10.1016/j.comnet.2005.01.008 |
Description (link): | http://www.elsevier.com/wps/find/journaldescription.cws_home/505606/description#description |
Published version: | http://dx.doi.org/10.1016/j.comnet.2005.01.008 |
Appears in Collections: | Applied Mathematics publications Aurora harvest 2 Environment Institute publications |
Files in This Item:
There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.