Study On Intrusion Detection Technology Based On Fuzzy Clustering

With the development of computer network technology and scale, people are confronted with more and more serious problems of information security. In a network environment, it is an important and focal problem of information security that how to find precisely and rapidly various intrusions in computer network. But in the face with rapid changed, updated network environment and various new attacks, the traditional Intrusion Detection System(IDS) are limited in the extensibility and adaptability. As a result, some advantaged technology such as artificial intelligence, machine learning and data mining are developed into intrusion detection to get good detection results.Based on the theory of data mining and intrusion detection, a fuzzy clustering based IDS model is presented in this paper. In order to get the uniform membership function in fuzzy clustering, the updating model for alterable membership function is introduced, and then an improved fuzzy clustering algorithm SFCM (Shift Fuzzy C Means) is presented, which can detect intrusion over network data streams in real time. The intrusion detection system FCIDS(Fuzzy Clustering based Intrusion Detection System) based on algorithm SFCM is designed and implemented. The architecture of FCIDS is introduced. The correctness and validity of FCIDS are identified through the experiments for the KDD CUP99 datasets and real-time network packets. The experiment results show that FCIDS is effective, and can notably improve detecting precision and decrease error-detecting.