摘要
为了提高P2P系统中资源定位效率并保证资源下载质量,提出了一种基于域推荐相似度的P2P信任模型(DRSTrust).该模型充分考虑了节点兴趣和身份对信任计算的影响,通过划分兴趣域,以解决节点间因兴趣不对称而难以建立直接信任关系的问题.在兴趣域的基础上,根据域推荐相似度加权个体的域推荐度来计算节点的域服务信誉,并利用域服务信誉来刻画节点在特定兴趣域内的服务行为细节.然后,提出了服务相似度的概念,通过使服务行为最为相似的节点成为邻居来激励好节点贡献资源,同时将恶意节点排斥在网络边缘,从而有效提高了P2P系统中节点信任计算准确度.分析及仿真结果表明,在迭代的收敛速度、下载成功率以及模型的安全性等方面,DRSTrust均优于现有的信任模型.
To improve the resource location efficiency and guarantee resource download quality in P2P networks, a domain recommendation similarity based P2P trust model （DRSTrust） is proposed. This model takes the impact of peer interest and identity on trust computation into consideration suffi- ciently. The problem that the direct trust relationship is difficult to construct due to the asymmetric interest is solved by using interest domain partition. Besides, domain recommendation similarity is used to weight each individual recommendation, so as to calculate the domain service trust of a peer, and then the service quality of a peer in each specific domain is evaluated based on the domain serv- ice trust value. Service trust similarity is also proposed to organize peers with similar service trust values together so as to give incentives to good peers and repel malicious peers to network edge, which can improve the calculation accuracy of peer reputation in P2P networks. The analysis and simulation results show that the convergence speed, the successful download ratio and the security aspects of the DRSTrust are superior to those of current typical methods.
出处
《东南大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2012年第5期803-807,共5页
Journal of Southeast University：Natural Science Edition
基金
国家自然科学基金资助项目(61103225)
江苏省自然科学基金资助项目(BK2010133)
关键词
信任模型
兴趣域
域推荐相似度
服务相似度
trust model
interest domain
domain recommendation similarity
service similarity
作者简介
邢长友（1982-），男，博士，讲师，changyouxing@126．com．
