Packet Routing in Dynamically Changing Networks: A Reinforcement Learning Approach

DMCA

Packet Routing in Dynamically Changing Networks: A Reinforcement Learning Approach (1994)

Cached

Download Links

Other Repositories/Bibliography

DBLP

by Justin A. Boyan , Michael L. Littman

Venue:	Advances in Neural Information Processing Systems 6
Citations:	227 - 3 self

BibTeX

@INPROCEEDINGS{Boyan94packetrouting,
    author = {Justin A. Boyan and Michael L. Littman},
    title = {Packet Routing in Dynamically Changing Networks: A Reinforcement Learning Approach},
    booktitle = {Advances in Neural Information Processing Systems 6},
    year = {1994},
    pages = {671--678},
    publisher = {Morgan Kaufmann}
}

OpenURL

Abstract

This paper describes the Q-routing algorithm for packet routing, in which a reinforcement learning module is embedded into each node of a switching network. Only local communication is used by each node to keep accurate statistics on which routing decisions lead to minimal delivery times. In simple experiments involving a 36-node, irregularly connected network, Q-routing proves superior to a nonadaptive algorithm based on precomputed shortest paths and is able to route efficiently even when critical aspects of the simulation, such as the network load, are allowed to vary dynamically. The paper concludes with a discussion of the tradeoff between discovering shortcuts and maintaining stable policies. 1 INTRODUCTION The field of reinforcement learning has grown dramatically over the past several years, but with the exception of backgammon [8, 2], has had few successful applications to large-scale, practical tasks. This paper demonstrates that the practical task of routing packets through...

Keyphrases

packet routing reinforcement learning approach dynamically changing network practical task past several year q-routing prof local communication nonadaptive algorithm simple experiment stable policy reinforcement learning precomputed shortest path q-routing algorithm network load minimal delivery time accurate statistic successful application reinforcement learning module critical aspect switching network

DMCA