基于不完全信息多阶段博弈的入侵路径预测

随着入侵的推进入侵者掌握的信息会逐步增加,依据新信息入侵者会找到更好的入侵路径并作出调整。为了使防御方能准确预测入侵路径,首先基于超图理论建立动态防御图并提出动态防御图更新方法,对入侵者的信息更新进行预测;然后建立不完全信息多阶段博弈模型对不同阶段入侵者的入侵路径调整进行预测;最后设计基于博弈的动态防御图路径预测算法,对完整的入侵路径进行预测。实验给出对入侵路径进行预测的典型实例,对实例结果的分析说明了模型的合理性与准确性。

Intrusion path prediction based on incomplete information multi-stage game

The target network information which intruder learn will gradually increase in the intrusion process. According to the new information , intruder will find a better intrusion path than before and adjust strategy. This paper presented a method which can more accurately predict intrusion path. First, it established a dynamic defense graph based on hypergraph theory and proposed the method to update dynamic defense graph. Second, it established incomplete information multi-stage game model. Finally, it designed the dynamic defense graph path predictive algorithm based on game. The experiment gives a concrete example about the model of predicting intrusion path. The reasonableness and accuracy of the model are illustrated by the analysis of the example results.