Scheduling Sporadic, Hard Real-Time Tasks with Resources

Abstract

A new dynamic voltage scaling (DVS) approach to schedule sporadic, hard real-time tasks with shared resource in a power saving way is proposed. Although DVS algorithms have been recognized as a feasible solution to save energy, the scheduling of such tasks has not been fully explored. Thus, in the paper, the problem of power-saving scheduling for sporadic tasks that share a set of serially reusable, single unit software resources is investigated. To evaluate the performance of DVSSR, a real application, Robotic Highway Safety Marker (RSM), is employed. In this application, DVSSR and other DVS algorithms are simulated and compared. Simulation results show that DVSSR offers reasonable trade-off between cost and power savings. In RSM application, DVSSR achieves 92.03% average power savings while comparing with the classic DVS. Keywords: hard real-time and sporadic task, resource-sharing, task scheduling, dynamic voltage scaling, power savings