Energy Efficient Mobile Computing :

Limited energy budget poses the ultimate constraint on continued use of the wireless mobile devices. Maximizing the effective usage of the hardware components is a solution to the problem of extending the battery lifetime. Because hardware components basically consume the battery energy at the end. However, software applications are the clients of the hardware and they, in fact, dictate the hardware component how and when to run. Thus, a well-managed co-ordination between hardware and software components leads to efficient use of the battery energy. In our research work, we are investigating this energy management problem from both OS and software application perspectives.

Mobile devices are equipped with hardware components that are capable of staying at different power consumption levels. If OS has enough knowledge about the applications running on it, it can put the h/w components in low power states as long as possible. Beside that applications can save energy by adapting energy-aware strategy. For example, an application that requires fetching and buffering of date, can save energy by adapting fixed buffering scheme or fixed time-interval scheme. An application may even offload a portion of its task to a remote server to save energy.

As ap part of this research thread, we proposed a energy consumption model which estimates the energy cost of an application running on a portable wireless device. The cost model has been validated by performing actual measurement of energy cost from experiments. This study helps us in improving the design of energy efficient applications for portable wireless devices. Moreover, we are to build a framework, where the OS and its applications collaborate to reduce system power consumptions.

Related Publications: