Improved Artificial Immune Based Intrusion Detection Model

With the widespread use of computer science and the rapid expansion of internet technology, the internet brought convenience to people’s daily life, but it also brought a few negative influence.The security problems it exposed to people caused great loss. While the traditional security and defense technology can not contral the complex and dynamic external environment, intrusion detection system (IDS), as the second network security parclose after the firewall, plays an important role. Therefore, how to develop a intrusion detection system with high efficiency and practical applicability has significant implications to the application and development of internet. The novel and active defense network intrusion detection technology becomes an important research challenge for the scholar domestic and overseas.Rely on its functional characteristics, that are immune recognition, immune coordination and immune response, biological immune system can prevent people from the intrusion of the foreign cells and pathogenes. It recognizes and analyses them, eventually eliminates them, which maintains the coordination and sustainability of living organisms. Meanwhile, the mainly responsibility of intrusion detection system is to detect the running status of network, and prevent it from the abnormal behavior of inside or outside, which guarantees the security and efficiency of network. It seems that there exists obvious similarity between the two from the functional mechanisms. Therefore, based on this similarity, scholar domestic and overseas designed quite a few artificial immune algorithms which were aplied to network intrusion detection system which has higher detectability and sovle the problem that traditional security and defense technology can not work out. This paper detailedly introduces the fundamental theory of intrusion detection technology and artificial immunology, researches the development and status of artificial immunology based intrusion detection technology, and eventually propose an improved artificial immune based intrusion detection model. Aiming at the limitations of existent network intrusion detection model with artificial immune theories, the novel artificial immune intrusion detection model improved the extended dynamic clonal selection algorithm proposed by Kim. For generating more qualified detectors, a novel gene recombinant algorithm was proposed. And the modified memory detector updating strategy is proposed to guarantee the activity of memory detectors. The experimental results show that the presented model can achieve better detecting rates and lower false detecting rate.