A Pattern matching based Network Intrusion Detection System

Intrusion detection system (IDS) has recently become a heated research topic due to its capability of preventing attacks from malicious network users. A pattern matching intrusion detection system has been proposed in this paper. The pattern matching based NIDS consists of four modules: collection module, analyze module, response module and attack rule library. We base this system on CIDF architecture. Realizing that string matching is the bottleneck, our system has improved the performance of detection engines due to an improved algorithm based on the current BM algorithm. Testing results demonstrate superior performance in terms of the detection speed of IDS