基于改进贝叶斯和风险评估的无线传感网络信任模型

Research on trust model of wireless sensor networks based on bayes and risk assessment

针对无线传感器网络加密手段和身份认证技术不能解决网络内部攻击问题,提出一种基于改进的贝叶斯和风险评估的无线传感器网络信任模型。该模型针对基本的贝叶斯信任模型进行改进,采用异常折扣因子防止非入侵因素导致的网络异常行为;为实现信任对时间的敏感性,提出了自适应遗忘因子来削弱过去行为的影响;为防止恶意节点的诋毁,利用偏离度来过滤某些推荐节点;同时提出了基于熵的风险评估机制,用于评价信任的不确定性。实验结果表明,所提出的模型具有较高的检测率与较低的误检率,能有效保障无线传感器网络安全性能,并延长网络的生命周期。

Encryption and authentication technology can not solve the internal attacks problem in wireless sensor networks.A new wireless sensor networks trust model is presented based on Bayesian and a risk assessment.The model is improved mainly based on Bayesian theory and risk evaluation mechanism,with the aim to prevent the abnormal network behavior caused non invasion reasons by introducing abnormal discount factor.To reflect the sensitivity of trust on time,an adaptive forgetting factor is introduced to weaken the past actions.To prevent malicious slander,the deviation is used to filter the recommended node.Since Bayesian theory does not have the function of evaluating the uncertainty of trust,the entropy based risk assessment mechanism is developed,and it can reflect the uncertainty of trust.Finally,through a large number of experiments,the results show that the proposed model has higher detection rate and lower false detection rate,with the advantages of ensuring network security performance and extending the network life cycle.