采用双异质群体演化博弈的网络安全防御决策方法

Network Security Defense Decision Making Method Based on Dual Heterogeneous Population Evolutionary Game Model

为提高网络安全防御决策的精准度,提出了采用双异质群体演化博弈的网络安全防御决策方法。针对网络攻防的竞争型决策特点,引入生物学种群异质概念,提出双异质群体演化博弈模型;将博弈群体对于博弈历史经验的反思机制模型化为泊松过程,改进经典复制动态方程,提出反思能力是提升模型决策速度的关键因素;结合网络安全防御单次决策的确定性需求和传统纳什均衡解的局限性,提出严格纳什均衡是双异质群体演化博弈稳定的充要条件,通过引入全局单调的势函数,给出了博弈模型存在严格纳什均衡的稳定性证明和仿真验证。理论分析和仿真结果表明,所提模型能够克服经典模型中50%概率的最优策略事实偏差,且当博弈群体的反思能力为1.5时,防御群体决策趋于稳定的速率提升了151%。

To improve the accuracy of network security defense decision,according to the characteristics of competitive decision and the limitations of traditional Nash equilibrium(NE),the heterogeneous population evolutionary game model and decision method of network security defense are proposed.Analyzing the difference in decision behavior standards between the attackers and defenders,and introducing the biological population concept to distinguish the two sides as different game populations,a dual heterogeneous population evolutionary game model of network security defense is constructed.Then an improved replicator dynamic equation is proposed based on the reflection mechanism.Considering the deterministic demand of network defense decision-making and the limitations of NE,introducing the potential function and designing the stabilized pure strategy,the optimal defense pure strategy selection algorithm is proposed.It is concluded that a strict NE is just the necessary and sufficient condition for dual heterogeneous population evolutionary game stability,and the stability proof and simulation analysis are carried out for the strict NE combining with the global monotone property of potential function.The theoretical analysis and simulations show that the proposed game model and decision-making method can ensure stable convergence of the defense strategy,and provide a practical and credible defense strategy for the precise decision-making of actual network attack and defense.