Abstract
Energy consumption on mobile devices has been studied with the objective of finding ways to extend the device's autonomy, responding to a clear requirement from the end user. Computation offloading is an energy aware design technique that moves the execution of software routines outside the mobile device. In this paper, we illustrate a case study of an energy aware image processing application for Android mobile systems, which exercises process offloading to describe a quantitative outline of its actual contribution to save energy in a mobile device. We compared the battery consumption of local and remote executions of selected benchmarks, finding that even though in some cases it is actually more economical to offload the process, the expenses of overhead and network usage have a high impact on the energy required by the application.