Abstract:In P2P distribution networks, seed-returning planning policy as one of key factors greatly influences system throughput and the bandwidth utilization. Aiming at enhancing the download efficiency and the system throughput, the directory server needs to schedule the connection process using the formerly collected information. Based on an analytical model, this paper presents a planning policy using the ant colony optimization approach, which takes the overall disjointedness of content among peers into account and therefore makes full use of network bandwidth resources. Experiment demonstrates that the proposed policy can maximize the disjointedness of connected nodes, reduce the average download time, enhance the overall system throughput and greatly improve the performance of P2P distribution network.

Seed-returning Planning in P2P Distribution Network Based on Ant Algorithm