基于区域资源聚集的P2P检索策略  

Searching strategy of peer-to-peer based on regional resource aggregation

作　　者：郑晓健[1] 李彤[2] 付铁威[3] 

机构地区：[1]昆明理工大学津桥学院计算机科学与电子信息技术系,云南昆明650106 [2]云南大学软件学院,云南昆明650091 [3]昆明理工大学计算中心,云南昆明650093

出　　处：《计算机工程与设计》2014年第11期3794-3798,3812,共6页Computer Engineering and Design

基　　金：云南省教育厅科学研究基金项目(2010C011);国家自然科学基金项目(61262024)

摘　　要：P2P网络规模的扩大导致基于洪泛的检索方法易产生严重的通信消耗,为此提出一种区域资源聚集模型和检索算法。对非结构化对等网络中分散的资源进行分层聚集,形成大粒度的资源实体,显著缩减网络规模;综合考虑影响资源检索命中率的多种因素,构造资源引用价值衰减函数以调节描述资源实体的引用价值向量和矩阵;检索时从区域资源簇中具有最大引用价值的资源组开始寻找所要的资源。实验结果表明,该方法有效控制了消息转发范围、检索命中率有显著提高。Expanding the scale of network P2P leads to the flooding-based search methods producing a serious communication cost problem. To solve the problem, a searching strategy and a model of peer-to-peer based on the regional resource aggregation were put forward. The distributed resources in unstructured peer-to-peer networks were aggregated hierarchically into the large granularity resource entity and the network size was reduced significantly. Considering multiple factors that affected resource re- trieval hits, a resource reference value decline function was presented which regulated a vector and matrix describing resource en- tity reference value. It began to search the resources with the maximum reference value resource group in the regional resource clusters and this process was gradually extended until it found so far. The simulations demonstrate it effectively limits the scope of the messages dissemination and improves the performance of the system.

关 键 词：P2P网络 资源聚集 引用价值衰减 区域资源簇 消息扩散 

分 类 号：TP393.01[自动化与计算机技术—计算机应用技术]