Energy aware mixed tasks scheduling in real-time systems

Abstract

We address the problem of scheduling mixed tasks which consist of n hard real time periodic tasks with shared resources and soft aperiodic tasks and consider two conflicting goals: reducing energy consumption and decreasing response time of aperiodic tasks. Firstly, we compute the optimal speed of periodic tasks and present a novel static mixed task scheduling (SMTS) algorithm. It schedules periodic tasks with the optimal speed and aperiodic tasks with maximum processor speed. Secondly, a dynamic mixed task scheduling (DMTS) algorithm which can reclaim dynamic slack time generated from periodic tasks and the constant bandwidth server to reduce energy consumption is proposed. In addition, it combines dynamic voltage scaling and dynamic power management techniques. Finally, we prove that the DMTS algorithm is feasible. The experimental results show that the DMTS algorithm reduces average 7.18% and 53.66% energy consumption × response time compared with SMTS algorithm and baseline algorithm, respectively.