基于信誉和风险评估的动态信任模型研究  

Research on Dynamic Trust Model Based on Reputation and Risk Evaluation

作　　者：王樱[1,2] 范双南 王静[2] 

机构地区：[1]湖南科技经贸职业学院,湖南衡阳421001 [2]衡阳师范学院计算机科学系

出　　处：《计算机时代》2011年第9期22-24,共3页Computer Era

基　　金：湖南省教育厅项目(10C0175);衡阳市科技局项目(2010KJ23)

摘　　要：网格环境中的信任问题是网格计算发展过程中必须解决的一个重要问题。网格节点间的行为信任具有动态性和不确定性。文章综合考虑了信任度评估的动态性和风险性,提出一种新的基于信誉和风险评估的动态信任模型。该模型引入惩罚机制、风险机制和推荐节点信任度更新机制,以评估信任关系的复杂性。仿真结果表明,该模型在有效抑制恶意节点上较已有模型有一定的改进。The trust in a grid environment is an important problem in the development of grid computing. The behavior trust among grid nodes is dynamic and uncertain. Comprehensively considering the uncertainty and risk of trust degree evaluation, we propose a novel dynamic trust model based on reputation and risk evaluation. The model introduces the mechanisms, including punishment, risk and updating the trust degree of recommenders, to evaluate the complexity of trust relations. The simulation results show that the model outperforms the existing models in restraining malicious nodes.

关 键 词：信誉 风险评估 动态信任模型 恶意节点 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]