Android malware detection based on permission and behavior analysis

The development of mobile Internet and application store accelerates the spread of malicious applications on smartphones, especially on Android platform. In this paper, we propose an integrated Android malware detection scheme, combining permission and behavior analysis. For APK files which had been detected, their MD5 values are extracted as signature for detection. For APK files which had not been detected, detection is carried based on permission and behavior analysis. Behavior analysis contained taint propagation analysis and semantic analysis. Experiment results show that this system can detect the malware of privacy stealing and malicious deduction successfully.