基于特殊数列的懒洪泛P2P搜索技术  

New Lazy Flooding Searching Techniques Based on Special Progressions in Unstructured P2P Networks

作　　者：孙晓莉[1] 禹继国[1] 

机构地区：[1]曲阜师范大学计算机科学学院,山东日照276826

出　　处：《计算机技术与发展》2010年第7期117-120,共4页Computer Technology and Development

基　　金：国家自然科学基金(10471078);山东省中青年科学家奖励基金(2005BS01016);山东省科技攻关计划项目(2009GG10001014);山东省教育科研项目(J07WH05)

摘　　要：非结构P2P网络中的搜索方法大都是直接地或者间接地对洪泛技术做了改进。全光网络中有三种懒洪泛技术,分别是门槛洪泛、指数洪泛和斐波那契洪泛。但是这三种技术是针对网络状态更新信息的分发,而不是资源搜索。把这一思想引入到P2P领域,提出了无结构P2P网络中基于特殊数列的懒洪泛搜索技术,方法是对资源的流行度进行分析以实现选择性的洪泛。不是将所有查询发给所有的邻居,而是只有资源的流行度到达一定值时才洪泛。这种技术大大提高了搜索的性能。仿真结果显示这种方法在搜索成功率、开销和延迟方面都有了较大的改进。Most of searching methods in unstructured P2P networks are all directly or indirectly improved from flooding technique.In all optical networks there are three lazy flooding methods：threshold flooding,exponential flooding and fibonacci flooding.However,the three methods are used to dissminate the information of updated network status,not used to search a resource.On the ground of the former work,put forward new lazy flooding techniques based on special progressions for unstructured P2P networks,it analyzes the popularity of a resource in order to decide when to flood.It floods only when the popularity of a resource reach a certain value.It significantly improves network performance due to selected flooding.Simulation results show that the method improves searching success rate,cost and delay.

关 键 词：无结构P2P 数列 搜索 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]