Research on Basic Problems of Cognitive Network Intrusion Prevention

Nowadays bad adaptive performance of network intrusion prevention system (IPS) becomes the bottleneck of its development in face of complex and fickle attacks. The idea of cognitive network intrusion prevention is creatively put forward to solve the problem in this paper. Three basic problems of the cognitive IPS, knowledge representation, environmental perception and self-learning, are discussed and some solving ideas of the problems are presented. Ontology-based knowledge presentation with meeting the need of semantic and the reasoning is presented for the first problem, The logical structure of security environmental perception is shown for the second problem, A new method of self-learning based on internal feedback from unexpected failure is proposed for the third problem.