Search method taking full consideration of both global and partial features in unstructured P2P networks
基于网络全局与搜索局部特性的P2P搜索算法

Most existing heuristic search methods only considered partial features existed in search process, or only paid attentions to the global features in networks. This paper put forward a search method called GLMW. The GLMW retained the advantage of dynamic topology adaptation in conforming to the global features in networks, and added consideration of the partial features in object searching process, optimizing current and subsequent requests according to historical search records and making better use of high capacity nodes to serve search process. The simulation results and analysis show that the GLMW can find more object replicas with lower delays, achieve better utilization of high capacity nodes.