A Distributed Reinforcement Learning Scheme for Network Routing
Carnegie Mellon UniversitySchool of Computer SciencePittsburgh, PA*also Cognitive Science Research Group, BellcoreMorristown, NJ
Abstract
In this paper we describe a self-adjusting algorithm for packet routing in which a reinforcement learning method is embedded into each node of a network. Only local information is used at each node to keep accurate statistics on which routing policies lead to minimal routing times. In simple experiments involving a 36-node irregularly-connected network, this learning approach proves superior to routing based on precomputed shortest paths.
1 Introduction
An algorithm which can monitor its own performance has the possibility ...
Get Proceedings of the International Workshop on Applications of Neural Networks to Telecommunications now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
