COMPACT DFA: A VARIABLE STRIDE PATTERN MATCHING ALGORITHM TO PERFORM PATTERN MATCHES USING HEXA

In any network identifying the intruders while packet transferring is done by using pattern matching. In every intrusion detection system different pattern matching approaches are used. One of the approach is construction of DFA to identify the exact pattern in the system. But memory usage and memory bandwidth are the bottleneck for the DFA construction. In this paper we propose an algorithm which identifies the pattern as variable strides i.e., it uses the block oriented approach instead of bit oriented process. It is a multiple pattern matching algorithm with minimum memory usage. With including the algorithm, we propose a compact DFA which does not use addition memory for traversing in the graph to identify the pattern. Using all these approaches the throughput of the system can be increased in many folds at minimum cost