Negative Selection Algorithm In Ids

With the development of computer network, illegal intrusion is increasing and the methods of attack are becoming more and more complex. It not only brings great economical loss to companies and individuals, but also affects national security and social stability. Intrusion detection technology is becoming an important component in the protection system of information security. In recent years, artificial immune has become a new research focus. Intrusion detection technology based on artificial immune is also becoming a research focus. And its main characteristic is simulating the basic principles of biological immune system, system framework and related algorithms to achieve the intrusion detection.This thesis deeply analyses the self-adaptability and scalability of intrusion detection system, which have inpacts on the efficiency of intrusion detection. The solutions are put forward to by deeply studying the negative selection algorithm and analyzing the weakness of the lack of adaptability and need of lots of self-elements in the algorithm.This thesis has mainly completed the following work:1. Because negative selection algorithm has some problems in the stage of production of detector, such as self-matching, the number of detectors, inflexible system framework and so on. It caused the overlapping of system space and reduced the detection efficiency greatly. To solve this problem, this thesis presents a new type of intrusion detection model. Feedback technique is used to adjust the self radius of self elements, the detection radius of detectors, the number of detectors. Then the coverage rate of system space and the detection rate are improved.2. Analysis of the basic principles based on the model of network intrusion detection are given. Moreever function and realization of each module is detaily elaborated.3. Experiments show that feedback negative selection algorithm performs well than negative selection algorithm in terms of detection rate, false alarm rate, the self-learning and adaptation.At last, the thesis summarizes the research work and comes up with ideas for further work.