社交化P2P系统中基于个体动态属性的共演化模型  

Dynamic Tag-based Coevolution Model for Social P2P Systems

作　　者：田琳琳[1] 李明楚[1] 金星[1] 王震 TIAN Lin-lin;LI Ming-chu;JINXing;WANG Zhen(School of Software, Dalian University of Technology, Dalian 116620, China;School of Cyberspace Security, Hangzhou Dianzi University, Hangzhou 310038, China)

机构地区：[1]大连理工大学软件学院,辽宁大连116620 [2]杭州电子科技大学网络空间安全学院,杭州310038

出　　处：《小型微型计算机系统》2018年第6期1250-1254,共5页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(61403059;61572095;61702076)资助

摘　　要：社交化P2P应用常利用空间互惠机制激励用户积极互动,分析网络群体中合作行为如何涌现成为自组织协作系统研究热点之一.为探讨个体特性与系统状态的协同演化现象,本文构建用户内在属性、行为策略与网络结构共演化模型,探讨基于动态标签的伙伴选择机制对合作演化的影响.作为社交网络中个体的代表性属性,年龄通常影响个体的行为方式及社交关系.本文度量个体间的关系强度时引入年龄相似度,并提出考虑系统年龄与策略年龄的结构演化模式.蒙特卡洛仿真实验结果表明,静态与动态年龄规则下这种断边重连方式形成利于利他个体的网络结构,其中策略年龄规则对网络群体中合作的促进作用最为明显.Social P2 P applications encourage cooperation through spatial reciprocity mechanisms,using social relationships of users.Many studies focus on how network group cooperative behavior emerges in such self-organization collaboration systems. To investigate the mechanism of partner selection based on dynamic tag,we construct a coevolution model of individual attributes,strategies and network structures. Age,as an intrinsic attribute of individuals in social networks,influences the social patterns of individuals. Considering system ages and strategy ages,an innovative pattern of network structure evolution mode is proposed and age similarity is introduced to measure the relationship strength between individuals. Simulation results show that such structure adjustment is beneficial to cooperation under static and dynamic age rules,and the rule of age strategy plays the most significant role in promoting cooperation among network groups.

关 键 词：演化博弈 伙伴选择 动态标签 社交化P2P 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]