基于位置感知与兴趣分组的混合式P2P架构研究-计算机软件与理论专业论文.docxVIP

RESEARCH ON HYBRID P2P ARCHITECTURE BASED ON LOCALITY-AWARENESS AND INTEREST GROUPING A Dissertation submitted in fulfillment of the requirements of the degree of MASTER OF ENGINEERING SCIENCE from Shandong University of Science and Technology by Sun Liting Supervisor: Associate Professor Xu Jianguo College of Information Science and Engineering June 2010 声 明 本人呈交给山东科技大学的这篇硕士学位论文，除了所列参考文献和世所公认的文 献外，全部是本人在导师指导下的研究成果。该论文资料尚没有呈交于其它任何学术机 关作鉴定。 硕士生签名： 日 期： AFFIRMATION I declare that this dissertation, submitted in fulfillment of the requirements for the award of Master of Engineering science in Shandong University of Science and Technology, is wholly my own work unless referenced of acknowledge. The document has not been submitted for qualification at any other academic institute. Signature: Date: 山东科技大学硕士学位论文摘要 山东科技大学硕士学位论文 摘要 摘 要 对等网络（Peer to Peer,P2P）近年来被大量应用在网络应用系统中。作为一种新兴 的网络计算模式，P2P 具有容错能力强、自组织、可扩散性好等优点。 P2P 模式突破了 传统的客户端/服务器模式，强调节点之间的“对等性”，即 P2P 系统中每一个参与节点 兼有服务器和客户端两种身份。其核心思想是通过参与节点之间的直接交互来实现资源 和服务的共享。但是，随着 P2P 网络用户量和规模的增加，P2P 环境下的信息量也随之 飞速增长，给用户在搜索、定位和获取资源上带来了极大的困难。 本文在针对提高 P2P 系统搜索效率的问题上进行了分析和研究。利用层次架构来实 现位置感知，提出按兴趣分组的混合式 P2P 架构模型 LAIP2P(Locality-Awared and Interest Grouping P2P )。讨论了三种算法，分别是节点加入算法、节点离开算法以及节点搜寻算 法。利用 NS2 来对 LAIP2P 进行仿真和模拟实验，并分别以节点与节点间延迟和查询的 hop 数来模拟本模型，仿真结果显示出依时间顺序加入的节点，在网络节点相当多的情 况下，本文提出的模型与 Gnutella 相比，提高了 P2P 系统搜索效率。 关键字：混合式对等网络架构，位置感知，兴趣分组 Abstract Peer-to-peer system (P2P system) is becoming a key technology and has been widely a platform for users to search and share file over the Internet. P2P as a new network compute mode, is self-organized and has good Expansibility. P2P break through the traditional Client/Sever model, and put stress on nodes parity. The core thought is use the direct interact o f participating nodes and realize the share of information. But the information amount rises rapidly with the intense increment of both users’quantity and network’s scale, and caused a serious problem in searching, locating and retrieving resour