Summary
International Symposium on Nonlinear Theory and its Applications
2009
Session Number:B1L-B
Session:
Number:B1L-B1
A parameter control method for Q-Routing for networks with breaking links
Shunsuke ARAKI, Masaharu ADACHI,
pp.-
Publication Date:2009/10/18
Online ISSN:2188-5079
Summary:
"In recent years, the scales of Wide Area Networks (WAN) and backbone networks have been growing because of increases in the number of network users. However, it becomes more difficult to manage traffic control in a centralized way as the scale of the network grows. It is difficult to adjust traffic in network environments that are dynamic (not only in terms of the network topology but also in terms of changes of network loading). Q-Routing is an autonomous distributed routing approach based on the framework of reinforcement learning, and its ability to adapt to changes of a network load is high. However, its ability to adapt to topology changes a network is low. In this paper, we propose a parameter control method for Q-Routing and show that it improves the ability of adapting to changes of network by breaking links."