Malware Variants Detection Using Behavior Destructive Features

Abstract

The variants of malware are a major threat to the security of computer systems. Millions of hosts on the Internet have been infected by malwares variants. Accurate detection of malware variants has become a key challenge for malware detection. The existing static detection is susceptible to file shelling and code obfuscation, while the dynamic detection is subject to anti-debugging and anti-virtual machine technology. Therefore, by combining the static and dynamic detection, we designed a malicious variants detection method based on behavior destructive features to analyze malicious samples.