Research And Design On Intrusion Detection System Based On Artificial Immune System

Apply the mechanism of human immune system to intrusion detection system and form intrusion detection system based on artificial immune system. It is proved by experiments that real-valued vector representation is better than binary representation in dealing with the complicated problem. The detection rate of detector set generated by random way is low. Propose two methods to build gene library and use gene library to generate detectors. The detector set generated by gene library is with higher detection rate. Reduce the times of comparing and enhance the detection speed by clustering non-self space into many sub-spaces and increase the detection rate by detecting the samples falling into"holes". The existing intrusion detection systems can detect some intrusions, but for other new type intrusions they may be vulnerable. Propose a dual-layer intrusion detection system model by studying the human immune system. It can enhance the protection level of intrusion detection system further.