A shadow price guided genetic algorithm for energy aware task scheduling on cloud computers

Abstract

Minimizing computing energy consumption has many benefits, such as environment protection, cost savings, etc. An important research problem is the energy aware task scheduling for cloud computing. For many diverse computers in a typical cloud computing system, great energy reduction can be achieved by smart optimization methods. The objective of energy aware task scheduling is to efficiently complete all assigned tasks to minimize energy consumption with various constraints. Genetic Algorithm (GA) is a popular and effective optimization algorithm. However, it is much slower than other traditional search algorithms such as heuristic algorithm. In this paper, we propose a shadow price guided algorithm (SGA) to improve the performance of energy aware task scheduling. Experiment results have shown that our energy aware task scheduling algorithm using the new SGA is more effective and faster than the standard GA. © 2011 Springer-Verlag.