Clustering-based anomaly detection for smartphone applications

Nowadays, Smartphones have been widely used due to their capabilities in communication and multimedia processing. Smartphones provide access to a tremendous amount of sensitive information related to business, such as customer contacts, financial data, and Intranet networks. Hence, the Internet of the future will be mobile Internet. However, threat of malicious software has become an important factor in the smartphones security. In this paper, a new behavior-based malware detection framework using three clustering methods (PAM, DBSCAN and t-distribution) is proposed. Experimental results show that the approach has high detection rate and low rate of false positive and false negative.