P2P环境下基于历史及推荐的信任模型  

History and recommendation-based trust model in P2P environment

作　　者：李雯[1] 谢冬青[1] 吴勇[1] 

机构地区：[1]湖南大学计算机与通信学院,长沙410082

出　　处：《计算机应用研究》2008年第3期915-919,共5页Application Research of Computers

基　　金：国家自然科学基金资助项目(60373085)

摘　　要：提出了一种新的peer-to-peer网络环境信任模型。明确区分了交易信誉和推荐信誉,不是简单地将交易信誉迭代作为推荐信誉,而是通过推荐偏移量来计算推荐信誉,并通过动态平衡自身历史经验和他人推荐信誉来计算节点的综合信任值,最后给出了模型的实现协议及仿真实验。实验结果表明,该模型计算更逼近真实值,且能抵抗恶意的联合攻击、动态策略型攻击等安全威胁,具有良好的性能。A new trust model was proposed in peer-to-peer networks. Peers＇ transaction reputation and recommendation reputation were definitely distinguished. Recommendation reputation was computed through the deviation of recommendation, not by the computation of iterative transaction reputation. And peer＇ s integrated trust value was computed dynamically by its history experiences and other＇ s recommendations. Finally, the model ＇ s implementation protocol and simulation were put out. The result is more close to the peers＇ true value, and this trust model can resist some security threats, such as malicious collusions and dynamic strategic attacks, with favorable performance.

关 键 词：对等网 信任 交易信誉 推荐信誉 推荐 安全 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]