Cloud Computing for Mobile Devices - Reducing Energy Consumption

Being powered by the batteries that are limited in their capacity is one of the main restrictions of the mobile devices. On the other hand, further enhancement of their characteristics and mobile Internet mounting speed incite the growth of user's demands. We want the most sophisticated applications to work rapidly and perfectly. Thus, we need a solution to improve the mobile devices' energy consumption. This master's thesis presents the approach which is able to decrease the power consumption on mobile gadgets. The core idea lies in porting the parts of the application's functionality to the remote server in order to relieve the CPU load. Consequently, we are going to derive the benefit from cloud computing. For this purpose, the " heavy-loaded " code blocks are extracted from mobile apps and built into the server-side applications. Both mobile and " remote " versions of the apps are run, their power consumption is measured and compared. It is expected that energy spent on client-server communication is less than power needed to execute the task on a phone or tablet. Moreover, the performance change is estimated. The experiments are conducted on three Android applications. The paper provides a technique to move the functionality to the cloud and to organize the communication between a client and a server. It is explained when and how it is required to adjust the approach to overcome the challenges that may occur. In addition, the restrictions and recommendations are denoted and performance changes are outlined. The applications are tested against various data volumes, Internet connection speed and use different ways to transfer data. Energy consumption for each case is measured and the results are presented.