Intrusion Detection System In The Mode Matching Algorithm

This paper illustrates the research background and significance, the development of intrusion detection and status, pattern matching algorithm development and current situation and overview of IDS, and focus on several pattern mating algorithms, including BF algorithm, KMP algorithm, BM algorithm, BMH algorithm, BMHS algorithm. It has detailedly analyzed the performance of these algorithms. Then to the deficiency of BMHS algorithm, and combining with advantages of BMH and BMHS algorithms, this paper has proposed two improved matching algorithm-BMHS2 and BMHS3 algorithm. We present the algorithm pseudo-code implementation and analyze the efficiency of the new algorithm. Finally we provide some improvement for these algorithms.The main work and contribution of this thesis:(1)The pattem matching algorithms of the intrusion detection are analyzed detailedly.lack of BMHS algorithm ale pointed out.(2)Combining the advantages of BMH algorithm and BMHS algorithm, a new pattern matching algorithm-BMHS2 was proposed. Based on BMH algorithm and BMHS algorithm, BMHS2 algorithm has increased the distance d calculations of P character appeared in the right position to P, and has had capabilities to skip many inefficient characters, to achieve the maximum moving distance of 2m. The time complexity in the best case is O(n/2m). Simulation results show that the improved algorithm is superior to the BMH algorithm and BMHS algorithm in the matching frequency and execution time, and can improve the detection efficiency of intrusion detection system.(3)To the deficiency of BMHS algorithm, this dissertation brings forward a new pattern matching algorithm-BMHS3 algorithm. Based on BMHS algorithm, BMHS3 algorithm increases last characters and a consideration of character presented in the preceding position of the matching string whether to appear. In the most cases, the maximum moving distance achieves to m + 1. The probability of maximum moving is higher than the resule from BMHS algorithm. Simulation results show that the improved algorithm is superior to the BMHS algorithm in the matching frequency and execution time, and can improve the detection efficiency of intrusion detection system.