Adaptive power management system for mobile multimedia device

Abstract

Recently, methods that combine the concepts of dynamic voltage scaling (DVS) and dynamic power management (DPM) approaches have been proposed for a single task. This study proposes adaptive power management system (APMS), which utilises useful features of both DPM and DVS and an existing pattern analysis algorithm, and new break even time-based task partition scheduling (BET-BTPS). APMS splits tasks based on BET to reduce the total power consumption of multiple multimedia applications in a mobile embedded system environment and analyses their usage patterns of peripheral devices and apply the results to their scheduling for further reduction of power consumption. It also takes into account the additional time delay that may occur while invoking and executing a timeslot scheduler to guarantee real-time services for multimedia data streams when the scheduling of APMS is in effect. In order to determine the optimal processor speed that minimises power consumption, the extra consumption required by the scheduler is compared with that reduced by BTPS. The proposed approach demonstrates better performance compared to existing power management schemes not only for single task but multiple tasks.