基于可信群划分及评估值波动性和一致性的可信评估模型  

A Trust Model Based on Volatility and Consistency in Trusted Groups

作　　者：殷安生[1] 张顺颐[1] 

机构地区：[1]南京邮电大学宽带无线通信技术教育部工程研究中心,江苏南京210003

出　　处：《南京邮电大学学报（自然科学版）》2014年第3期101-105,共5页Journal of Nanjing University of Posts and Telecommunications：Natural Science Edition

基　　金：江苏省科技成果转化专项资金(BA2012047);南京邮电大学人文社科项目(NYS213026)资助项目

摘　　要：提出一种基于兴趣相似度的可信群划分策略。将全局评估转化为群内评估,既解决了评估消息集聚的问题,又区分了不同兴趣域内的评估结果,提高了系统的服务成功率。同时提出一种结合评估值波动性和一致性的权重计算方法。仿真结果表明,新模型能抵抗各类恶意节点,且比以往的模型服务成功率更高。A classifying strategy for trusted groups is presented according to the interest similarity.The global trusted evaluation is transformed into the local evaluation in the trusted group,thus it addresses the centralizing problem of evaluation news and separates evaluation results among different interest groups to improve the success of serving for the system.Meanwhile,a weighted calculating method is proposed by combining the volatility with the consistency of the evaluation value.Simulation results show that the proposed model can combat all kinds of malicious nodes,and yield higher serving success rate than the previous models.

关 键 词：可信群 兴趣相似度 权重 波动性 一致性 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]