Artificial Immunization Technology In Intrusion Detection Applied Research

Along with information technology's high speed development, the network security problem based on the immunity principle's intrusion detection system's research already became focus.This article first elaborated the intrusion detection system's concept, the development present situation and the trend of development, the intrusion detection model and the classification, and has analyzed the existing intrusion detection system existence question, then introduced immune system's structure, the immunity mechanism.Existing system model existence false negative and false positive excessively high question, and was not high to the known intrusion's variation and the unknown intrusion detection rate already became an important question that influence performance. In view of such situation, this article proposed one kind of new pattern matching algorithm and the improvement detector production algorithm. Because the traditional pattern matching algorithm has the r value difficulty, and had not considered that the different character position to matches the result creates the influence, This article proposed one kind of improvement's pattern matching algorithm, this matching algorithm based on the gene fragment's principle, had considered the different position character section weight different factor, causes the pattern matching to be more reasonable fast; The detector was an intrusion detection system's core, detector's efficiency has decided the intrusion detection system's detection rate directly. The tradition uses the negative selection algorithm the detector, because has the big redundancy, causes the intrusion detection system's detection rate drop. In view of the above question, this article proposed one kind of new intrusion detection model, in the research existing detector production algorithm's foundation, comprehensive oneself has the algorithm superiority to make the improvement to the negative selection algorithm, gives one kind of more highly effective detector production algorithm.Finally proved through the experiment, examines the attack type and the number using the improvement algorithm is friends with in the traditional algorithm, false negative and false positive are also low.