基于概率和可信度方法的普适计算信任模型

Probability and credibility-based trust model for pervasive computing

根据普适计算环境下信任的特点以及现有信任模型存在的不足,提出一种基于概率和可信度方法的信任模型。模型综合考虑影响信任的主观和客观因素,通过主观信任、直接信任和推荐信任得到总体信任。计算推荐信任时引入基于可信度的匹配度算法对多个推荐实体进行取舍,在计算前就把不符合推荐条件的实体排除,再结合对推荐链路的更新,能够更有效地避免恶意推荐的影响,计算出来的信任值可靠度更高。

According to the traits of trust in pervasive computing and the shortages of the existing trust model, a probability and credibility-based trust model was given. Considering the subjective and objective factors, the trust was acquired by subjective trust, directive trust and recommendatory trust. Based on the credibility-based matching degree algorithm, the entities falling short of recommendatory condition were excluded before calculation. And combining the update of recommendatory chain, the problem of vicious recommendation was solved. The trust was more credible.