Research On The Immunity-Based Model Of Network Intrusion Detection System

Internet not only brings covenience to our live, but also has a lot of unexpected problems, computer security is a prominent issue. As the important component of current network security architecture, the network intrusion detection system is to distinguish between normal behaviour and the abnormal behaviour mainly. Human immune system protects body against pathogenic organisms from the outside through identification of self and non-self. Immune system is a highly robustness, adaptability and autonomous system. It provides reference to us to construct more reliable intrusion detection system. We build an immunity-based network intrusion detection system base on the research on the artificial immunization, algorithms and models.Negative selection algorithm and clonal selection algorithm has been designed to make it applicable to the network intrusion detection system. Several detectors'generated algorithms have been analysed. The immune-based intrusion detection system uses the r-chunks match rule. A single detectors and the set of detectors has been analysed their performance through templates. And r-chunks match rule has been anaysed the holes compare to the r-continuous bits match rule.The system has a lot of detection nodes, and they constitute a distributed artificial immunity-based intrusion detection system. Each detection node has some capabilities to identify non-self. And nodes can communicate each other to share detective capacity. In the immune-based intrusion detection system the mathematics ways is designed to describe the system problem. The premise has been given to design a distributed, extensible and robustness system.