基于Markov微分博弈的移动目标防御决策优化

随着网络攻防向实时连续和动态高频变化的方向发展，传统的离散多阶段网络攻防博弈模型已难以满足实际需求，而且传统网络攻防模型中的节点状态单一，难以准确描述实际网络对抗中节点类型的演化过程。将节点传染病动力学模型加以改进并应用到网络攻防对抗中，用来描述攻防过程中不同状态节点的演化过程及节点状态间的迁移关系。在构建移动目标Markov微分博弈防御模型时，各阶段内运用微分博弈模型分析，阶段间运用Markov决策过程描述状态转移，通过均衡分析和求解，设计防御决策优化算法。最后，通过仿真实验验证该模型和优化策略的可行性和有效性。

Research on defense decision optimization of mobile targets based on Markov differential game

With the development of network attack and defense towards real-time continuous, dynamic and high-frequency changes, the traditional discrete multi-stage network attack and defense game model has been difficult to meet the actual needs, and the node state in the traditional network is single, which is difficult to accurately describe the evolution process of the node type in the actual network confrontation. This paper improved the dynamics model of node infectious disease, and applied it to network attack and defense. The model described the evolution process of nodes in different states and the migration relationship between nodes in the process of attack and defense. When constructing the Markov differential game defense model for moving targets, the paper used the differential game model to analysis in each stage, and used the Markov decision process to describe the state transition between stages. Through equilibrium analysis and solution, the paper designed the defense decision optimization algorithm to analyze the node state evolution. Simulation results show that the feasibility and effectiveness of the proposed model and optimization strategy are efficient.