The Mixed Intrusion Detection System Based On Immunology

Because the problem faced in computer system is similar with immune system: immune system protects body from the harm of antigen and the computer security system protects computer form intrusion, thus, it is more significant to put the methods and principles of biology immune into the fields of computer safety system.This article first elaborated the intrusion detection system and its relative technologies, then discussed the biological immunity system principle, the immunity system composition, the immune cell function, the immunity process of the immunity system, the feasibility and necessary of the network intrusion detection system based on the immunity principle, and the basic concept of artificial immunization system. We analyzed some key technologies on emphasis, such as the differentiation of the Self and Non-self set, matching rules and detection algorithms. The new NA matching rule was proposed and the algorithm was presented, involving negative selection and clonal selection. Some experiments were designed to demonstrate the validity of the new rule and algorithm, too.Finally, this article has produced an intrusion detection system model based on immunology. On the experimental platform we test the model with KDDCup99 empirical datum Through comparion, we can see that the immunology model has higher detection rate and better real time, and can find the new intrusion.