摘要： 根据网络攻击具有并发性,攻击特征的提取具有不确定性等特点,给出了采用模糊Petri网实现攻击知识的表达和入侵检测的推理模型。该模型解决了误用入侵检测系统中现有知识表示方法不能并行推理的问题,以及传统的基于Petri网可达图搜索求解导致模型描述复杂、推理缺少智能的问题。最后通过入侵实例验证了该模型的正确性和有效性。

关键词: 入侵检测, 模糊Petri网, 知识表示, 不确定推理

Abstract: In terms of the concurrence of the network intrusion and the uncertainty of attack character extraction, fuzzy Petri nets to express attack knowledge and construct reasoning model are given. This model can reason parallelly to detect the concurrent attack, and avoid the complexity of traditional Petri net reasoning based on searching reachability set for the solution. The reasoning has some intelligence. The experiment proves the validity and efficiency.

中图分类号: 

• TP393.08