Detection of unknown applications in smartphones: A signal processing perspective

Different applications in smartphones result in different power consumption patterns. The fact that every application has been coded to perform certain tasks leads to the claim that every action on-board (whether software or hardware) will consequently have a trace in the power consumption of the smartphone. However, similar power consumption patterns are observed for the same application irrespective of the types of operating system they are executed on. Therefore, it is safe to further claim that no two applications can have similar power consumption patterns as it is highly unlikely that they have exactly the same source code. The idea behind this work is that by analyzing only the power consumption signals of a smartphone, valuable information regarding its operation can be revealed. In view of this, the authors propose a systematic and generic methodology that involves performing some signal processing techniques on the power consumption signals of different applications in smartphones to detect and separate unknown activities/applications from known ones. In this work, we present a proof of concept supported with some test cases to validate the approach. The preliminary results hold promise in detecting malware in smartphones.