The Research Of Intrusion Detection System Based On Pattern Matching

With the popularization and fast development, network users face increasingly serious security issues, thus network intrusion has become the most important threat to the computer security and network security. So network intrusion detection system(NIDS) appears as the keystone and hotspot in the computer security research field which emerge as the times require.This dissertation analyses and summarizes the current status of intrusion detection research, focuses on research and practice on pattern matching algorithm which is technique difficulties in network intrusion detection ,proposes some problems intrusion detection technology must face and research trend.First, this dissertation summarizes model, constitutes, category and trend of IDS, and then pattern matching which is applied widely is discussed in detail from the aspect of theory and technology principle and performance of mostly pattern match algorithm such as BF, KMP and BM. This paper bring two forth modified algorithms for BM. These improved algorithms reduce comparison number and improve time performance contrasted with the standard BM algorithm. The paper also brings a new pattern matching method which is based on the frequency of character appearing.The paper also discusses a new intrusion detection technology: protocol analysis. Through contrasted of pattern match and protocol analysis, It gets a conclusion that integrating these two methods into NIDS is necessary for current NIDS.At last, this dissertation analyses some problems intrusion detection technical must face and discusses expectation of IDS in future.