dc.description.abstract |
Nowadays, user needs for mobile solutions are growing up rapidly, because of the rapidly grow of needs for speed, entertainment and mobility at wireless communication. Manufacturers are facing huge challenges to make devices with enough power efficiency that cover these needs for energy. To improve the energy-efficiency of mobile devices, the engineers need appropriate tools. One method to estimate energy consumption of smartphone devices (i.e. android phones) is Energy Profiling (for example, using reference implementation - Android Power Profiles). This method allows the energy consumption estimation online, i.e. without using any external devices, while using reference data obtained via prior using of offline measurements tools. This is a good method for application developers to improve their applications to be faster with less energy consumption. Another method is the method of offline measurements which usually done by using external measuring device with a “reference” device for testing. It is even recommended to measure with a “fake” test battery, which is just source of direct current with fixed level of voltage. This approach helps to minimize interference of battery properties on measured values.
This thesis describes the steps to build a functional infrastructure to collect and process the performance and power figures and builds a power model based on the gathered measurements. The implementation will target Android based platform. |
el |