P2P网络中基于全局信任值的激励机制  

Incentive Mechanism Based on Global Trust Values in P2P Networks

作　　者：欧阳竟成[1,2] 林亚平[3] 周四望[3] 李文[1,2] 

机构地区：[1]湖南理工学院信息与通信工程学院,岳阳414006 [2]湖南省普通高等学校复杂系统优化与控制重点实验室,岳阳414006 [3]湖南大学信息科学与工程学院,长沙410082

出　　处：《系统仿真学报》2013年第5期1046-1052,共7页Journal of System Simulation

基　　金：国家自然科学基金(60973127);湖南省教育厅科研项目(10C0759;07A024);湖南省自然科学基金(10JJ2044);湖南省高校科技创新团队支持计划资助(湘教通[2012]318号)

摘　　要：针对自由骑问题,提出一种基于全局信任值的激励机制,以激励节点积极贡献、可靠服务。在大规模的P2P网络中,节点间交易的随机性造成建立重复博弈的概率极低,导致TfT(Tit-for-Tat)策略失效。节点的全局信任值能够反映节点对网络贡献大小与服务质量好坏,基于节点间全局信任值的差异设置服务响应规则、查询转发规则以及邻居选择规则,形成一种普适性更强的激励机制,间接体现TfT策略。仿真实验表明,这种激励机制能够抑制自由骑,鼓励节点共享资源。Aiming at the flee-tiding problem, an incentive mechanism based on the global trust value of nodes was proposed to encourage nodes to contribute actively and service dependably. The randomicity of trades between nodes results in the probability to build up a repeated game is very low, and brings about the TIT （Tit-for-Tat） strategy to loss effect in the large-scale P2P network. Global trust values of nodes can reflect their contribution and service quality to networks. Based on the difference of global trust values between nodes, some rules including response of service, forward of query and selection of neighbor are established to form an incentive mechanism which has good universality and reflect the TfT strategy indirectly. Simulations show the incentive mechanism can suppress free-riders and encourage nodes to share resources.

关 键 词：对等网络 激励机制 自由骑 全局信任值 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]