P2P信任系统中基于重复博弈的惩罚机制研究  

RESEARCH ON PUNISHMENT MECHANISM IN P2P TRUST SYSTEM BASED ON REPEATED GAME

作　　者：杨先伟[1] 战学秋[1] 

机构地区：[1]无锡职业技术学院,江苏无锡214121

出　　处：《计算机应用与软件》2017年第6期328-333,共6页Computer Applications and Software

基　　金：国家自然科学基金资助项目(11471144)

摘　　要：P2P网络是一种动态的、自组织的、分布式的开放网络环境,信任问题在P2P系统中扮演了越来越关键的角色,而惩罚机制的设计又是P2P信任问题的关键所在。设计了一种基于重复博弈的惩罚模型,创新点在于:在博弈论的分析框架之下,引入了基于信任度的惩罚策略并创造性地设计了惩罚度转换函数,使得在单次博弈中无法实现的信任和合作在重复博弈中得以实现。该惩罚机制具有良好的威慑性、容错性、主观性和区分性,并通过模拟实验证明了其可行性和有效性。P2P network is a dynamic, self-organized, distributed open network environment. Trust plays anincreasingly important role in P2P systems, and the design of penalty mechanism is the key of P2P trust. This paper designs a penalty model based on repeated games. The innovation is that in the framework of the game theory analysis, the introduction of the punishment strategy based on the degree of Trust and creatively designed a penalty conversion function, so that trust and cooperation can not be achieved in a single game can be achieved in the repeated game. The penalty mechanism has a good deterrence, fault tolerance, subjectivity and differentiation. The simulation experiment proves its feasibility and effectiveness.

关 键 词：惩罚机制 P2P 信任 博弈论 重复博弈 信任度 

分 类 号：TP31[自动化与计算机技术—计算机软件与理论]