基于链路状态质量的连续证据序列的信任模型研究

Trust Model with Continuous Behavior Evidences Based on Link State Quality

基于证据理论的信任模型是IP分布式网络中最常用的信任模型.在实际应用中需要解决两个问题:一是如何在缺少可靠的网络上层证据或者无法获取网络上层证据情况下计算信任值;二是在使用底层证据作为信任计算基础时必须考虑底层证据有连续稳定的特点.本文通过定义链路状态质量评价函数和对证据的信任态度函数,改进关于连续行为的链路状态证据特征的可信值计算方法,提出了在IP网络中使用连续链路层证据评价机制来考查节点的信任关系的模型.最后模拟无线IP网络环境中进行测试,结果验证表明:基于链路状态质量评价的连续证据序列的可信值计算,在上层信任证据缺乏的情况下可作为上层证据评估的很好的补充和纠错,使得信任计算结果更为合理和可信.

Trust model based on evidence theory is a most commonly used trust model in IP distributed network.There are two problems that need to be solved. One is the calculation of trust values with inadequate and unreliable upper layer evidences,the other being the continuous and stable characteristics of the link layer ＇s evidences. A new method is proposed in this paper. The three points included are defining evidence distrust function link state evaluation function and trust attitude function for evidence,calculating continuous credible evidence,and proposing trust model based on continuous link layer evidences in IP networks. Finally,experiments are carried out to simulate a wireless IP network. The experiments show that the method can be more credible to express trust value in case of lacking confidence evidences from upper layers.