Research On Detector Algorithm Under Neighborhood Space In Immune Intrusion Detection System

Along the rapid development of network technology, more and more peoplepay close attention to to computer on network security, Traditional intrusiondetection technology can not adapt to a variety of network attacks. Immuneintrusion detection has attracted a growing number of researchers. as a veryeffective intrusion detection technology.The Artificial Immune System(AIS) is anew computation theory inspired by biological immune system. The perfactcharacteristic of AIS with robustness, distributivity and self-organization makesIntrusion Detection System(IDS) based on AIS has become a hot topic in theresearch of network security.This article describes the analysis of the classification of intrusiondetection,and the develop status of intrusion detection research domestic andforeign. After comparing the mainstream of intrusion detection algorithm, selectneighborhood space as the space of the intrusion detection research. In thereseach of neighborhood negative selection algorithm, this paper focus onenhancing adaptive of the neighborhood space, improve the detectionefficiency.Due to clonal selection algorithm for limited antibody populationdiversity and premature problems, clonal selection algorithm is proposed to buildthe environment to improve the algorithm in neighborhood space.For the issues raised above, by analyzing the characteristics of theneighborhood space, we use a discrete method based on fuzzy cluster to improveneighborhood negative selection algorithm. It use neighborhood intrusiondetection immune improved clonal selection algorithm. Neighborhood clonalselection algorithm introduce to new cloning operator Maintained balanc betweenantibody promotion and suppression, out of local search,and through acombination of high frequency mutation operator and recombinantantibodies.Algorithm is increasing ability of the global and local search. For the problem of the negative selection algorithm with neighborhood is notflexible enough, a new algorithm is proposed called neighborhood negativeselection algorithm base on fuzzy cluster. The algorithm combines discretemethod based on the fuzzy cluster, the premise of not destroy data, the algorithmdivides dynamically the continuous data, building neighborhood shape space.Experimental results show that the negative selection algorithm based on thefuzzy cluster can effectively build the neighborhood shape space,control the sizeof the shape space and enhances the adaptive algorithm.By studying the algorithms described in this article, the adaptive of theneighborhood negative selection algorithm and the detection rate can beeffectively improved. Neighborhood clonal selection algorithm robustness, fastconvergence time, the time complexity is not high.