Android phone based appraisal of app behavior on cell networks

Abstract

The rapid adoption of smartphones has engendered a large ecosystem of mobile data applications. In fact, a large part of mobile tra c is now data and not voice. Many of these applications, for example VoIP clients, stay active in the background. In the background, they may not communi- cate large amounts of data. However, their regular bursts of activity can lead to large signaling overheads, wastage of radio resources, and draining of a phone's battery. Signal- ing overheads can lead to service outage over a large geo- graphical region by overloading the radio network controller (gateway) of the 3G (LTE) network, which is expected to handle signaling from a large number of base stations. In this work we propose for Android smartphones an on-the- phone mechanism to detect background applications that due to bad design (given the network's settings) or their malicious nature (exploiting the network's settings) lead to above mentioned ine ciencies. Speci cally, we propose the metrics of average energy/byte and the average time-to-state- promotion (TSP) after the Radio Resource Control (RRC) enters the IDLE state. The metrics capture an application's e ciency, which is a function of the network's settings of RRC inactivity timeout values and an application's back- ground activity. The e cacy of these metrics is tested on commonly used Android applications. We also outline a fully functional ready-to-install tool that we developed and used for our studies.