TARSB：一种基于相对搜索介数的P2P拓扑演化策略

近几年来,对等网络(P2P networks)得到了很好的发展,然而其巨大的网络开销也逐渐成为阻碍其进一步推广的主要问题.在其最重要的应用领域--P2P文件共享系统中,早期的系统往往基于一种单一的拓扑结构,忽略了成员之间的差异对网络本身的影响,使得其理想的结构与实际网络之间存在很大差距,难以达到预期的搜索效率.针对P2P文件共享系统进行研究,提出了相对搜索介数(Relative Search Betweenness)来刻画各种异构性对搜索过程的综合影响,并设计了一种基于相时搜索介数的拓扑演化策略:TARSB.该策略通过动态自适应地调整覆盖网络结构来平滑各结点在搜索性能上的差异,从而达到提高网络整体搜索性能的效果.实验表明,TARSB策略在保证搜索命中率的情况下,大大降低了搜索报文的数量,并使网络自主演化成一种更加稳定和高效的拓扑结构.

TARSB:Topology Adaptation Based on Relative Search Betweenness in P2P Networks

Peer-to-Peer(P2P)applications become very popular in current networks,but the large traffic consuming limits their further spread.In the field of P2P file-sharing system,one of the most important applications of P2P networks,early systems usually obtain a single and flat topology,which ignores the heterogeneity of peers.To give a comprehensive measure of heterogeneity in P2P file-sharing systems,especially in the searching phase,we defined a parameter,called Relative Search Betweenness(RSB).We also proposed...