Application of MAS in Implementing Rational IP Routers on the Priced Internet

Pricing and prioritizing the Internet access can be one of the potential measures to improve the efficiency of the Internet utilization and control the Internet congestion. In order to meet a diversity of user access requirements, the IP (Internet Protocol) routers can provide users with various priority service classes. This paper explores the application of multi-agent systems (MAS) in the implementation of rational IP routers on the priced Internet. The negotiation model for task decomposition and allocation, and the payment model are described. The strategies of achieving a more optimal global-level resource allocation and a rational and fair local-level resource allocation and scheduling for tasks with different priority are proposed. The strategies for price setting and adjustment are discussed. The consumer strategy is also addressed.