International Symposium on Nonlinear Theory and its Applications


Session Number:A2L-B



Control of Selfish Routing Based on Replicator Dynamics with State-Dependent Tax

Takurou Misaka,  Takafumi Kanazawa,  Toshimitsu Ushio,  


Publication Date:2009/10/18

Online ISSN:2188-5079


PDF download (291.7KB)

Selfish behaviors of users can increase average data transmission latency in computer networks. A selfish routing game is a simple model of the selfish behaviors and its replicator dynamics has been proposed. To reduce the inefficiencies due to the selfish behaviors, a control method to impose latency on each transmission path as a state-dependent tax has been proposed. Several properties of the control method have been discussed using the model based on the replicator dynamics. In this paper, we propose a routing algorithm based on replicator dynamics modeling an effect of the state-dependent tax. Moreover, we show the effectiveness of the proposed algorithm using a network simulator.