移动无线传感网中恶意软件传播的最优安全策略  

Optimal Security Strategy for Malware Propagation in Mobile Wireless Sensor Networks

作　　者：曹玉林[1,2] 王小明[1] 何早波 

机构地区：[1]陕西师范大学计算机科学学院,陕西西安710062 [2]青海师范大学计算机学院,青海西宁810008

出　　处：《电子学报》2016年第8期1851-1857,共7页Acta Electronica Sinica

基　　金：国家自然科学基金(No.61373083;No.61363080)

摘　　要：移动无线传感器网络的大规模应用依赖于建立起应对恶意软件攻击的安全策略.一个有效的防护措施就是对传感器节点安装免疫补丁或清除节点中的病毒.考虑到传感器节点的移动特性,根据传染病学理论我们建立了恶意软件传播的动力学模型.基于此模型提出了以易感节点免疫比例与感染节点恢复比例作为优化控制变量的最优目标函数,使得在任意终止时刻被感染的节点数量最少并且实施安全措施成本最小.通过平衡点的稳定性分析,得到了恶意软件传播与否的阈值.运用庞德里亚金(Pontryagin)极大值原理得到了免疫比例与恢复比例的最优控制变量对.仿真结果表明,该模型对于建立遏制恶意程序在移动无线传感器网络中扩散传播的安全策略具有指导意义.It is imperative to take reliable measures against outbreaks of malware to enable large-scale application of mobile wireless sensor networks （MWSNs）.An effective measure is through dissemination of security patches to immunize or heal the nodes.Considering the mobile behavior of nodes,a malware propagation dynamic model is proposed based on ep-idemic theories.We formulate an optimal objective function subject to this model with immunization and heal proportion as optimal control variables.Our goal is to minimize the cost of the two measures and the number of infected nodes.Through stability analysis of equilibrium,we find the threshold whether a malware propagates or dies out.We used Pontryagin’s max-imum principle to obtain the optimal levels of the two controls.Simulation results show that the model play a significant role in building up the strategies of preventing malware from propagation in MWSNs.

关 键 词：最优控制对 传染病模型 基本再生数 庞德里亚金极大值原理 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]