Static Analysis Based Behavioral API for Malware Detection using Markov Chain

Abstract

Researchers employ behavior based malware detection models that depend on API tracking and analyzing features to identify suspected PE applications. Those malware behavior models become more efficient than the signature based malware detection systems for detecting unknown malwares. This is because a simple polymorphic or metamorphic malware can defeat signature based detection systems easily. The growing number of computer malwares and the detection of malware have been the concern for security researchers for a large period of time. The use of logic formulae to model the malware behaviors is one of the most encouraging recent developments in malware research, which provides alternatives to classic virus detection methods. To address the limitation of traditional AVs, we proposed a virus detection system based on extracting Application Program Interface (API) calls from virus behaviors. The proposed research uses static analysis of behavior-based detection mechanism without executing of software to detect viruses at user mod by using Markov Chain.

Keywords: Malware Detection; Markov Chain; Virus Behavior; API Calls