新颖的面向网络服务的动态信任模型

A novel dynamic trust model for network service

提出一种新颖的面向网络服务的动态信任模型。引入信任评价与区间直觉模糊数的对应关系,给出基于信任需求属性的经验信任、邻居推荐信任及代理推荐信任的计算、合成、及综合运算。同时,采用精确函数比来确定信任属性的权重因子,而时间序列权重因子则由指数分布的逆形式确定。最终以总体信任的得分函数及心性因素来区分信任服务。仿真实验表明,该模型可有效的建立实体间的信任关系并遏制恶意实体的攻击。

A novel dynamic trust model for network service is proposed.The corresponding relationship of trust evaluation and interval-valued intuitionistic fuzzy numbers are introduced in the computations,compounds,and general operations of experience of trust,neighbor recommendation trust and agency recommendation trust,which are based on trust demand attributes.Meanwhile,the trust attributes’ weight factors are determined by precise function proportion,while time series weight factors are defined by the inverse form of exponential distribution.And the trust service is finally distinguished by the result of general trust scoring function and psychological factors.Simulation results show that the model can effectively build trust relationship between entities and restrain attacks from malicious entities.