目前缺乏有效的方法对已存的各种信任模型进行分析和评估。为解决该问题,提出了一种基于可信建模过程的信任模型评估算法。将信任模型按照信任生命周期分解成信任产生、信任建模、信任计算、信任决策和信任传递这5个部分,对每个过程进行可信性分析,并模糊量化评估值,用Bayesian融合形成综合的评估结果。最后以无线传感器网络信任模型为例,对所提算法进行了验证,同时给出了算法的有效性分析和仿真。结果证明所提算法能解决给定信任模型的最优化选择问题,并在准确性、稳定性和效率3个方面相对以往算法有一定的改善。该评估结果为信任模型的标准化提供了一定的参考。
There are few effective evaluation methods for assessing trust models. This paper presents a trust model evaluation algorithm where the trust model is divided into five trusted processes in the trust lifestyle such as trust establishing, trust modeling, trust computing, trust decision-making and trust transmissions. Then, each process is analyzed for reliability using fuzzy theory and the Bayesian method with the results fused into a comprehensive quantitative evaluation result. The method is illustrated by evaluating wireless sensor network trust models. The analysis results show that the algorithm can provide optimal selection of trust models in practical applications with better accuracy, stability and efficiency than prevous methods. The evaluation results provide a reference for the development of standard trust models.