Open Access Open Access  Restricted Access Subscription Access

Power Consumption Based Android Malware Detection


Affiliations
1 School of Computer Science and Technology, Civil Aviation University of China, No. 2898, Jinbei Road, Tianjin 300300, China
 

In order to solve the problem that Android platform's sand-box mechanism prevents security protection software from accessing effective information to detect malware, this paper proposes a malicious software detection method based on power consumption. Firstly, the mobile battery consumption status information was obtained, and the Gaussian mixture model (GMM) was built by using Mel frequency cepstral coefficients (MFCC). Then, the GMM was used to analyze power consumption; malicious software can be classified and detected through classification processing. Experiment results demonstrate that the function of an application and its power consumption have a close relationship, and our method can detect some typical malicious application software accurately.
User
Notifications
Font Size

Abstract Views: 60

PDF Views: 0




  • Power Consumption Based Android Malware Detection

Abstract Views: 60  |  PDF Views: 0

Authors

Hongyu Yang
School of Computer Science and Technology, Civil Aviation University of China, No. 2898, Jinbei Road, Tianjin 300300, China
Ruiwen Tang
School of Computer Science and Technology, Civil Aviation University of China, No. 2898, Jinbei Road, Tianjin 300300, China

Abstract


In order to solve the problem that Android platform's sand-box mechanism prevents security protection software from accessing effective information to detect malware, this paper proposes a malicious software detection method based on power consumption. Firstly, the mobile battery consumption status information was obtained, and the Gaussian mixture model (GMM) was built by using Mel frequency cepstral coefficients (MFCC). Then, the GMM was used to analyze power consumption; malicious software can be classified and detected through classification processing. Experiment results demonstrate that the function of an application and its power consumption have a close relationship, and our method can detect some typical malicious application software accurately.