对基于CAN的P2P网络模型的改进研究  

Improved Research of P2P Model of CAN

作　　者：谷科[1] 危韧勇[1] 

机构地区：[1]中南大学信息科学与工程学院,湖南长沙410075

出　　处：《小型微型计算机系统》2007年第3期416-420,共5页Journal of Chinese Computer Systems

基　　金：湖南省自然科学基金项目(03JJY3111)资助.

摘　　要：提出了一种对CAN(Content-Addressable Network)的改进方法.该方法基于CAN模型拓扑结构,在资源分布中采用了二叉树形式进行资源配置优化,并且在路由策略中加入了向后学习算法,从而使得改进后的CAN模型更加有效.实验是利用C++语言和网络仿真工具NS2来模拟整个模型在实时中的运行状况.实验结果表明改进后的模型提高了资源定位的效率和容错性,使得整个网络资源分布更为合理,路由性能显著提高,在非结构化的P2P网络模型中值得推广应用.This paper proposes an improved method of CAN （Content-Addressable Network）. with the architecture of CAN , the method exploits the architecture of binary tree to make resources＇ configuration become reasonable in the distribution, and adopts backward-studying algorithm in routing algorithm. So the improved CAN becomes more effective. Experiments are imi- tated to perform by exploiting C＋＋ and net-imitated tools （NS2）. Experiment results indicate that the improved CAN improves resource orientation and error solution, and makes net resources＇ distribution become more reasonable and improves routing function. So it is worth popularizing and applying in unstructured P2P network.

关 键 词：点对点 哈希冲突 向后学习 资源定位 路由算法 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]