Intrusion detection based on immune clonal selection algorithms

Abstract

Immune clone selection algorithm is a new intelligent algorithm which can effectively overcome the prematurity and slow convergence speed of traditional evolution algorithm because of the clonal selection strategy and clonal mutation strategy. We apply the immune clonal selection algorithm to the process of modeling normal behavior. We compare our algorithm with the algorithm which applies the genetic algorithm to intrusion detection and applies the negative selection algorithm of the artificial immune system to intrusion detection in the dataset kddcup99. The experiment results have shown that the rule set obtained by our algorithm can detect unknown attack behavior effectively and have higher detection rate and lower false positive rate. © Springer-Verlag Berlin Heidelberg 2004.