受恶意程序传染的WSNs可生存性评估

Survivability Evaluation for WSNs under Malware Infection

为实现传感节点数据高可靠传输的无线传感器网络WSNs(Wireless Sensor Networks),提出一种针对受恶意程序传染的聚簇WSNs可生存性评估方法。通过使用能预测恶意程序传染行为的博弈模型,将恶意程序传染的故意性与马尔可夫链的随机性关联,实现利用马尔可夫链中的状态转换描述恶意程序传染行为的目的。使用提出的易感节点可靠度评估指标,根据可靠性理论将聚簇WSNs看作一个"并-串-并"系统,推理得到了一个簇、一条路由直至整个WSNs的可靠度评估指标,最终得到WSNs生存期评估指标。实验分析了影响WSNs可生存性的因素,结果表明提出的方法能有效评估WSNs的可生存性,为设计高可生存的WSNs提供理论基础。

To ensure dependable transmission of sensed data from sensor nodes in Wireless Sensor Networks （WSNs）,we propose a measurement of survivability evaluation for WSNs under malware infection. By introducing a game model that can predict malware behavior,we relate the intent of malware infection to a Markov chain（MC）for describing a successful infection as a MC state transition. We then set up a novel measurement representing the reliability of a susceptible sensor node. After considering clustered WSNs as a parallel-serial-parallel system,we deduce reliability measurements for a cluster,a route,and the entire WSNs. Finally,we attain the measurement to calculate the survival time of the entire WSNs. Experiments show the influence of parameters on the survivability of the entire WSNs,and validate the effectiveness of the proposed method that can build up a theoretical foundation for highly survival WSNs.