Intrusion Detection Based on Simulated Annealing and Fuzzy C-means Clustering

Abstract

An intrusion detection method based on simulated annealing and fuzzy c-means clustering is proposed against the problems of sensitivity to initialization and local optimal solution caused by fuzzy c-means clustering algorithm. The ability of simulated annealing algorithm jumping out of the local optimal solution combined with fuzzy c-means clustering is firstly used in order to get global optimal clustering, and normal and anomaly data are identified by normal cluster ratio. Then the identified clusters can be used in the detection of intruding action. The experiment in the KDDCUP99 data set indicates that the method has a better detecting effect than traditional fuzzy c-means algorithm.