基于Gnutella协议与划分技术的P2P网络模型的设计与实现  

P2P Network Model's Design and Implementation Based on Gnutella Protocol and Division Technology

作　　者：庄雷[1] 郭永强[2] 潘春建[1] 

机构地区：[1]郑州大学信息工程学院 [2]河南电力研究院

出　　处：《计算机应用研究》2004年第9期253-255,共3页Application Research of Computers

基　　金：国家"863"计划资助项目(2001AA11114)

摘　　要：Peer-to-Peer是一种很有发展潜力的网络技术,在对等计算、协同工作、搜索引擎、文件交换等领域有很好的应用前景。在传统的P2P网络模型中,网络上的所有主机不管它们所处的距离远近,它们的地位都是一样的,这在实际应用中却行不通。针对这种情况,在新的模型中将网络划分为不同的子网,子网内与子网外的主机分别对待。在子网内,采用IP多播技术自动定位主机节点;在子网外,则采用手工输入和自动发现相结合的方法来定位主机节点。在查询机制中,也对内部网络和外部网络采用了不同的查询方法并用Java实现了该模型。Peer-to-Peer(P2P) is a type of net work technology that will develop greatly in the future.It broad applies in the field of equal computation,coordinatedwork,searched engine and file exchanged.In the traditional P2P network model,all the peers in network are equivalent without reference to their positions,but it's not true in practice.For this reason,the whole network is divided into different small areas and use different method to deal with peers.Inside the small network,we use IP multicast technology to find a peer automatically.Outside the small network,we find peers by both automatically finding mechanism and manually entering a host list method.And in the query mechanism,we also use the different methods.Finally,we make out software based on the new P2P network model.

关 键 词：GNUTELLA 泛滥式查询 短路效应 IP多播 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]