Energy-Efficient Tasks scheduling Algorithm for Dual-core Real-time Systems

Abstract

Real-time embedded devices have been widely used in our daily life. To satisfy the performance requirements, most current designs tend to apply the dual- or multi-core processor architecture in the systems. Such systems, however, usually have low power consumption demands. Therefore the Dynamic Voltage Scaling (DVS) technique has been included in most designs. In this paper, we focus our study on the energy-efficient task scheduling algorithm for the dual-core real-time systems. Our goal is to minimize the system’s energy consumption and maintain the performance of task execution at the same time. To achieve this goal, we propose two approaches: off-line and on-line. For the off-line approach, we propose an Integer Linear Programming (ILP) based algorithm to find the optimal scheduling. For the on-line approach, we propose a heuristic algorithm. The experimental results show that the energy consumption can be reduced effectively by the heuristic algorithm, and is close to the optimal bounds obtained by the ILP model.