A novel peer selection strategy in P2P VoD system using biased gossip

Abstract

The Peer-to-Peer (P2P) streaming media service based on video-on-demand is increasingly being acquainted and used with present rapid growth of Internet. Adopting reasonable peer selection mechanism to guarantee improved QoE (Quality of Experience) is one of the hot issues in the research of P2P streaming media system. However, most of the recent studies focus on peer selection by using centralized server, and a simplex peer selection mechanism is used. It is difficult for the network established by these methods to provide efficient data transmission without considering the heterogeneity and complexity. This paper, presents a novel peer selection algorithm which considers each performance function value of a peer as basis of a neighborhood establishment. In addition, a modified gossip protocol is adopted to periodically exchange partial network view information among peers and adjust peer selection at the next moment, adaptively and dynamically.