非结构化P2P网络中基于节点的MQR算法设计与实现  

Design and Implementation of Node-based MQR Alogorithm in Unstructured P2P Networks

作　　者：谢晃 张昱[1] 王云凯[2] 

机构地区：[1]中国科学技术大学软件学院,江苏苏州215123 [2]西南财经大学经济信息工程学院,成都611130

出　　处：《计算机工程》2014年第9期111-116,123,共7页Computer Engineering

摘　　要：在非结构化P2P搜索中,由于缺少全局性的管理机制,网络节点无法获得整个网络的拓扑结构及目标数据的定位信息,因此查询消息的路由过程具有较高的随机性,不仅查询性能低,而且宽带消耗大。为在有效控制网络冗余消息规模的同时提高数据的搜索范围,在分析现有2类典型非结构化P2P路由算法的基础上,提出一种基于节点的MQR算法。利用网络节点的状态信息及搜索过程中查询消息的TTL值状态信息,从数据的搜索范围与网络使用情况2个方面来提高非结构化P2P网络搜索性能。仿真实验结果表明,与传统的P2P路由算法APS和Random Walk相比,该算法在搜索准确率、网络利用率及召回率方面有更好的表现。Due to the lack of global governance mechanisms in the unstructured Peer-to-Peer（P2P）network,network nodes do not know the entire network topology and target data location information. So the query message routing process has a high randomness,not only query performance is low,but also bandwidth consumption is large. Based upon the analysis of two typical categories of unstructured P2 P routing alogorithms,this paper proposes a node-based Mixed Query Routing（MQR）alogorithm to deal with the scale problem of redundant messages and to improve the search scope of data. By means of the status information about the nodes and the TTL values of the queries,it can improve the search performance both in the aspect of data＇s search scope and network efficiency. Simulation experimental results show that compared with the typical alogorithms APS and Random Walk,the MQR alogorithm can reach higher accuracy rate,better network efficiency and recall rate.

关 键 词：对等网络 资源定位 路由算法 非结构化 MQR算法 

分 类 号：TP393.02[自动化与计算机技术—计算机应用技术]