Optimal Server Selection and Pricing in Receiver-Driven P2P Streaming Systems

In receiver-driven P2P streaming architectures, a client peer may get resources from one or more server peers. The optimal peer selection problem is to minimize the total streaming cost subject to the bandwidth requirement. However, P2P peers are usually selfish identities who seek to maximize their own benefits. In this paper, we analyze the optimal peer selection problem in a game theoretic approach and use the well-known VCG procedure to schedule the pricing policy. We inspirit the true revelation of server peers´ cost value in order to minimize the total streaming cost. Our contributions also include a study of practical implementation of VCG and a discussion of the pricing mechanism in a repeated-game view.