个体理性受限的P2P网络合作激励研究

项目名称： 个体理性受限的P2P网络合作激励研究

项目编号： No.61202425

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 计算机科学学科

项目作者： 王锐

作者单位： 北京航空航天大学

项目金额： 23万元

中文摘要： 节点间相互提供资源和服务是P2P 网络的原动力，因此，合作激励是P2P 网络研究中的关键问题。近年来P2P 网络的应用场景逐步拓展到了无线网络、在线社会网络等新领域，在这些场景中面临着节点行为随时间和环境变化、噪声干扰、信息受限等问题，此时经典博弈论已不适用，给合作激励的研究带来了新的挑战。 本项目针对上述问题，将P2P 网络节点看作具有有限理性的博弈参与者（即博弈参与者 具有获取最大收益的预期，但是在计算、存储、信息获取等方面受到一定约束），在P2P 网 络节点有限理性的表达与博弈建模、基于自适应博弈学习的P2P 网络合作激励方法、合作激 励方法的抗噪音扰动、信息受限条件下的合作激励四个方面开展研究。通过本项目的实施， 促进对P2P 网络节点行为特征的深入理解，为合作激励问题在新应用场景中面临的挑战提供 新的解决方案。研究成果还可以为以经济学最新研究成果解决计算机网络问题提供新的思 路。

中文关键词： 对等网络；社会网络；有限理性；博弈论；

英文摘要： Sharing resources and services among different nodes as the basis, P2P networks rely on cooperation between nodes greatly. Therefore, cooperation incentive policies play very important role in P2P networks. In recent years, the application of P2P scheme extends to wireless networks and online social networks. In these areas, the cooperation incentives have more challenges since the nodes of the networks may change with time and different environments, or the communication between nodes may be interfered by noises, or even the network nodes have only limited information about the environment. In this project, we consider the P2P network nodes as bounded rationality game players, meaning that the nodes only have limited computing, storage capability and limited information. These assumptions can describe the application more realistic, but we cannot apply the results of previous researches based on classic game theory, which holds only with the assumption of completely rationality. We will carry out the following research works: 1) expression of bounded rationality of nodes in P2P networks and bounded rational game modeling; 2) cooperation incentives with adaptive game learning in P2P networks; 3) anti-noises solution in cooperation incentives; 4) cooperation incentives with limited information. Through the res

英文关键词： Peer-to-Peer；Social network；Limited Rationality；Game Theory；