Investigating clustering algorithm DBSCAN to self select locations for power based malicious code detection on smartphones

There is a growing number of viruses, malware, and other threats designed to gain access to system resources and information stored on smartphones. Previous work has shown that combining naive location identification with power-based malicious code detection can successfully detect certain kinds of power-hungry malware. This paper will discuss the effectiveness of using clustering, a more sophisticated manner to identify locations, compared to the previous naive approach. The paper will also discuss the methodology used to collect more accurate location data while avoiding a noticeable impact on the phone's battery life.