面向非对称认知的移动目标防御动态博弈方法

Dynamic game method for asymmetric cognition oriented moving target defense

移动目标防御(MTD)通过主动地、动态地“移动”系统安全配置,扰乱攻击方对系统状态的认知,进而影响攻防决策和演化过程,被称为“改变游戏规则”的网络空间革命性技术.MTD的技术本质是建立和控制攻防双方的非对称认知.为了实现对这种非对称认知关系的量化分析,通过建模攻防知识面在攻防双方认知中的差异化投影,将非对称认知关系引入到MTD博弈决策中,提出面向非对称认知的MTD动态博弈模型和基于多阶段潜在风险分析的决策方法.数学实例分析和模拟实验验证了模型和方法不仅能实现面向非对称认知的攻防博弈推演,也进一步提高了攻防状态建模的灵活度和空间规模,对基于博弈的网络安全管理具有指导意义.

Moving Target Defense(MTD)actively and dynamically“moves”the security configuration of the system to disrupt the attacker’s cognition of the system,thereby affecting the attack-defense decision-making and evolution process.MTD is called“game-changing”cyberspace revolutionary technology.The technical essence of MTD is to establish and control the asymmetric cognition of the attacker and defender.In order to realize the quantitative analysis of the asymmetric cognition relationship,this paper introduces the asymmetric cognitive relationship into the MTD game decision-making by modeling the differential projection of the attack-defense knowledge surface in the cognition of the attacker and defender,and proposes an asymmetric cognition-oriented MTD dynamic game model and a multi-stage potential risk analysis based decision method.Mathematical case analysis and simulation experiments verify the effectiveness of the model and method,which not only realize the ability of attack-defense game deduction based on asymmetric cognition,but also make the modeling of attack-defense states more abundant and flexible.The model in this paper has guiding significance for game theory based network security management.