Adaptive and Efficient Peer Selection in Peer-to-Peer Streaming Networks

Nowadays Peer-to-Peer technologies for content distribution such as file-sharing and streaming media delivery are quite popular with Internet users. For P2P streaming applications, an appropriate peer selection scheme is critical to reduce startup delay and to provide media delivery of high quality. In this work, we propose a novel peer selection scheme AEPS that is both efficient and adaptive to dynamic P2P network. The AEPS combines the advantages of round-trip time based (RTT-based) schemes and available bandwidth based (ABW-based) schemes, to perform efficient and accurate selection. Experiments under three distinct scenarios are realized to quantitatively compare the AEPS with other selection schemes. The experiment results exhibit that the AEPS is more accurate than RTT-based schemes and more efficient than ABW-based schemes, achieving impressive performance in general.