Paper Info
Title
Evolutionary Learning in Computational Ecologies: An Application to Adaptive, Distributed Routing in Communication Networks
Abstract
We outline the characteristics of the adaptive, distributed routing problem in communication networks and discuss the problem in relationship to recent research in distributed artificial intelligence and computational ecologies. We offer a brief overview of genetics-based machine learning applied to fuzzy control and present details of a Pittsburgh-style fuzzy classifier system which is used as a routing control agent. Experiments are described in which distributed fuzzy routing controllers are evolved in a small-scale and symmetrical, simulated network. The performance of evolved fuzzy controllers is compared to that of traditional controllers using the shortest-path routing algorithm. Possible ways forward in extending this approach to routing control in real networks are suggested.
Year
DOI
Venue
1995
10.1007/3-540-60469-3_29
Evolutionary Computing, AISB Workshop
Keywords
Field
DocType
communication networks,evolutionary learning,computational ecologies,fuzzy control
Intelligent control,Neuro-fuzzy,Computational intelligence,Computer science,Fuzzy routing,Fuzzy logic,Artificial intelligence,Computational learning theory,Fuzzy control system,Machine learning,Fuzzy rule
Conference
ISBN
Citations 
PageRank 
3-540-60469-3
0
0.34
References 
Authors
9
3
Name
Order
Citations
PageRank
Brian Carse125926.31
T C Fogarty21147152.53
Alistair Munro316618.26