Research Of Intrusion Detection Technology Based On Fuzzy Clustering Analysis

Intrusion detection technology as a kind of positive safe protection technology effectively compensate for the lack of traditional security protection mechanism. However, with the diversity of intrusion techniques,traditional intrusion detection technologies cannot meet the needs of current network environment.Applying artificial intelligent technology to intrusion detection technology has been a focus among recent researches.Cluster analysis is an unsupervised learning techniques. The clustering analysis applied to intrusion detection can extract the knowledge and rules with the potential value in the data.and find abnormal behavior. Because the object data mining is the mass, many new high-dimensional data clustering algorithm was proposed, such as the clustering algorithm based on model, clustering algorithm based on density and fuzzy clustering algorithm, etc. Intrusion detection technology based on the cluster analysis improved the ability of processing massive data and detection capabilities, and makes the intrusion detection system has to self-learning, self-organizing ability.Based on the research of background ,the main work of this paper are as follows:1. Considering data noise, error factors can affect the fuzzy C-means clustering algorithm,a kind of outlier identification method based on the density was proposed in this paper.This method overcomed the defect that the traditional fuzzy C-means clustering algorithm is more sensitive to outlier.2. Due to genetic algorithm exist deceptive problem and immature convergence problem in the evolutionary process, a kind of fuzzy C-means algorithm based on the improved gennetic algorithm is proposed after outlier identification which improve the global search ability, overcomes that the original algorithm is more sensitive to initial value and the local optimal solution is insufficient.3. KDDCUPl999 data set by the data simulation experiment, the experimental results show that the improved algorithms in intrusion detection have good detection performance, can obtain higher detection rate and lower the rate of false positives.