【作者】 陈志琦；

【导师】 苏德富；

【作者基本信息】 广西大学 ， 计算机应用技术， 2005， 硕士

【摘要】 P2P网络是一种自组织、没有中央职权和基础设施的分布式系统,因为其参与的自发性和自治性,使P2P网络中资源的可用性有着极高的变数和不可预知性。且实际应用中,由于大多数自主的参与者因缺乏激励而不共享资源,传统的P2P系统(比如Napster、Gnutella等)广泛地出现Free-Riding问题,严重阻碍了P2P技术和应用的发展。由此引发了许多关于P2P激励机制方面的理论研究课题和商业计划。本文在对近年来使用经济学模型分配分布式系统资源的相关研究进行分析比较的基础上,借用经济学中的博弈均衡理论,提出一种基于服务质量差异的激励模型来提高P2P系统性能和效率。 本课题研究的主要内容包括以下几个方面: 一、研究P2P网络底层架构,以Gnutella为原型,引入对等网专用设备(Peer Server),为激励模型提供良好的底层网络平台。 二、构建对等网文件共享系统的激励机制。这一部分首先分析了P2P网络中参与者之间对网络资源的竞争关系,并在非合作博弈Nash均衡理论框架下构造一个基于服务质量(QoS)差异的P2P激励模型。然后在改进后对等网络底层模型上,通过仿真实验分析激励机制下不同性质的用户采取不同策略时的均衡状态。 三、借用分布式微观经济理论,将网络资源系统看成是一个价格随供求关系浮动的竞争市场,网络用户需要购买网络资源以满足个人的服务质量(QoS)。利用Flow Control技术模拟网络资源的分配优化过程,给出其仿真试验的测试结果及评价。更多还原

【Abstract】 Peer-To-Peer (P2P) networks are self-organizing, distributed systems, with no centralized authority or infrastructure. Because of the voluntary participation, the availability of resources in a P2P system can be highly variable and unpredictable. In particular, there is a free-rider problem in traditional peer-to-peer networks such as Napster: individual users are provided with no incentive for sharing their own files and thereby adding value to the network. As a result, a large number of academic and commercial projects are under way to develop P2P systems for various incentive mechanisms. In this paper, we use ideas from Game Theory to study the interaction of peers, and propose a differential service-based incentive scheme to improve the system’s performance.In this paper, by analyzing the disadvantages of pure P2P Network------Gnutella, a new routing policy is proposed, which can effectively improve its scalability and reduce its redundancy in the transmission of information.Base on this Network model, the incentive issues that arise in such file sharing systems will be addressed. We construct a formal game theoretic model of the system and analyze equilibrium of user strategies under several situations.Finally, we give the result of the simulation of the incentive model using a distributed microeconomic flow control technique, which models the network as competitive markets. In these markets resource owners price their resources based on supply and demand, and users purchase resources so as to maximize their individual Quality of Service (QoS).更多还原