HSAG-FC: Hybrid of Simulated Annealing with Genetic algorithm and Fuzzy Classification for IDS

Abstract

Nowadays signature attacks are termed as very big problem because it leads to software vulnerability.Malware writers confuse their malicious code to malicious code detectors such as Signature-based detection. However,it fails to detect new malware. This research article addresses the signature based intrusion detection from IntrusionDetection (IDS) systems. The proposed hybrid techniques for Generation of Signature are done using GeneticAlgorithm (GA) and Simulated Annealing (SA) approaches. For this, signature-set in execution statements are selected by using simulated annealing and genetic algorithm, which produce the optimal solution of selection. Then thegenerated signatures are matched with IDS by using the two pattern matching techniques, namely (i). Finite stateautomaton based search for Single Pattern matching technique and (ii) Rabin Karp string search algorithm for multiple pattern matching technique. These techniques are used to match the signature as in an effective manner. In addition tothis the Fuzzy Logic classification is used to find the degrees of truth of vulnerability for classification. The aim ofthe proposed work is to improve the final resultant accuracy in compared to existing techniques. The proposed RabinKarp- fuzzy logic system returns the higher performance metrics namely precision is 88% and Recall is 80% and inopen source dataset it contains 30 vulnerabilities this proposed worked well in detecting 28 vulnerabilities/ defect, theaccuracy of this proposed is 94.27%.