Energy Aware Proportionate Slack Management Scheduling for Multiprocessor Systems

Abstract

Multiple processors have been incorporated on the autonomous mobile robots in order to assure the higher performance needs of the real time applications. These robots are battery powered systems usually executes periodic real time tasks. The main challenge here is to professionally manage the energy consumption of the system to exploit the battery life time. Whenever a task terminates the implementation prior to its worst-case implementation time then slack has been established in the processor. It is the focal factor for increasing the energy consumption of the system. Numerous methods have proposed to handle the processor slack. Allocation of slack fully for reclamation or procrastination doesn't provide energy efficient loom. This article presents Energy aware proportionate Slack Management Scheduling (EAPSM) mechanism. It efficiently distributes the slack between sluggish and suspension to maximize the energy economy. It also reduces task migration and frequency switching overheads of the system. Simulation experiments illustrate that proposed algorithm reduces energy consumption of the system comparing with the existing optimal schedulers.