Date Added: Apr 2013
This paper presents a novel scheme to derive power saving profiles based on the usage patterns of the Android devices. The entire architecture is developed as an Android app "Power Monitor v2" and is deployed to the smart devices. A monitoring module of the app periodically collects several data from the devices and stores them locally. A learning engine then operates on the raw data to generate multiple usage patterns over time and space, which characterizes the user contexts. The engine then processes the patterns to generate power saving profiles dynamically within the devices. The profiles contain several system settings of the smart devices and intelligently optimize power consumption.