Energy aware fixed priority scheduling for real time sporadic task with task synchronization

Abstract

Abstract Energy management is one of the key issues in the real-time embedded systems. The slowdown method based on the dynamic voltage scaling (DVS) and the shutdown method based on the dynamic power management (DPM) can effectively reduce energy consumption. The previous algorithm which uses the static speed to deal with the blocking situation is too conservative. We present a static task synchronization for sporadic tasks scheduling (STSST) algorithm. The dynamic speed is computed to minimize energy consumption while satisfying the time constraints in the STSST algorithm. But, the STSST algorithm assumes that each task executes with its worst case execution time. As the actual execution time of the task is often lower than its worst case execution time. A dynamic task synchronization for sporadic task (DTSST) algorithm which can reclaim the slack time generated from the early completion task is proposed. It combines the DVS technique and the DPM technique to reduce energy consumption. Furthermore, the feasibility conditions are given and proved. The experimental results show that the DTSST algorithm can reduce energy consumption up to 20.57%∼67.66% over the STSST algorithm and it consumes 43.79%∼67.66% less energy than that of the DS algorithm.