Energy Efficient Location Tracking and Computation Offloading for Mobile Devices

In today's mobile devices, the battery reservoir remains severely limited in capacity, making energy consumption a key concern in the design and implementation of mobile applications. In this thesis, we closely examine two approaches to improve the energy efficiency of mobile applications: smartly utilizing the local computation resources and adaptively offloading the computation to the remote cloud. We use location tracking as a case study for the former. We identify the defects of conventional location tracking services that rely only on GPS, and develop SensTrack, a novel solution that jointly use use multiple sensor hints in modern smartphones. For the latter, we measure the energy reduction of computation offloading for two realworld mobile applications and identify the key influential factors. We then formulate the energy-efficient scheduling problem for computation offloading and present effective solutions.