Energy-Aware Scheduling Tasks on Chip Multiprocessor

Abstract

This paper presents a two-stage energy-aware schedule strategy to solve the problem of effective scheduling task set with interdependence on chip multiprocessor (CMP) system while saving energy. We propose a tri-dimensional coding based hybrid genetic algorithm (TCH) to allocate the task set on processor cores to minimize the execution time. Compared with traditional genetic algorithm, TCH uses the tri-dimensional coding scheme to match the architectural characteristics of CMP. We represent a heuristic algorithm using dynamic voltage scaling (DVS) method to reduce the system energy consumption. The simulation experimental results demonstrate that the strategy proposed in this paper efficiently allocates and schedules the tasks on CMP system.