The Clonal Selection Theory In Intrusion Detection Applications

Along with the computer network daily popularization and theapplication, the network information security also more and more isvalued people's. The first generation of safekeeping of security technology uses perimeter defense mechanism and so on the authentication, authorization, access control, firewall and encryption resists the invasion behavior, but regarding the complex network system, these passive defense measures already lacked the ability to do what one would like day after day.The invasion examination took the second-generation network security technology, at present received domestic and foreign has widely taken.Because the magnanimous network data multiplicity, not definite and the invasion method unceasing renewal, causes to the network data codeand its characteristic analysis method adapts the realistic demand with difficulty, the invasion examination has some to wait for the solution question. In recent years, the emerging intelligence computation technology - immunity algorithm which proposed based on the biosphere immunity system principle gradually is becoming the new fundamental research hot spot. As the immunity algorithm one basicalgorithm, the clone choice algorithm has received very many person of attention. This article in the research immunity principle and in the clone selective principle foundation, has made the improvement to the existing clone choice algorithm.This article depends on the Sichuan science and technology hall basic research project "the intelligent invasion examination system essential engineering research", the main work to invaded the examination, the biological immunity system principle and the work mechanism from the overall has carried on the introduction; This article proposed a dynamic clone choice algorithm and the immunity network unify new invasion examination algorithm; And carried on themodelling to the algorithm efficiency to confirm.In to the original algorithm improvement in, joins the gene bank and the high frequency variation the dynamic clone choice algorithm,enhances the detector the evolution speed. The immunity network andthe dynamic clone choice algorithm union, eliminates the redundancy which in the mature detector exists, enhances the detector the production efficiency. The memory detector production joins the variation and the competition to the algorithm in, the promotion memory detector evolution, enhances the community the affinity level.In the modelling, has used the American Columbia University computerscience is invades the network invasion test collection which the examination research group provides to the system prototype to carry on the confirmation, the system simulation experiment result hadproven that, after the improvement algorithm had the enhancement inthe convergence rate and the examination efficiency.