A Vulnerability Exploits Detection Method Based on Binary Program Behavior Analysis

Nowadays, Vulnerability exploiting has become a major means of malicious code communication. In order to solve the problem that polymorphic and metamorphic vulnerability exploit variants is difficult to detect, it presents a method of detecting vulnerability exploits based on binary behavior analysis in this paper. The method track and monitor the memory and register changes the binary file's behavior results, and have a formal verification on application's behavior through intermediate‐level language. In this paper, we examine the performance of binary analysis by taking two sets of experiments for separate targets, one is 13 local file vulnerabilities, the other is web browser vulnerabilities. The results show that the exploits detection method based on binary behavior analysis can be effectively used for analysis and detection of the vulnerabilities, also with significantly reduced time and space complexity.