Cooperative versus non-cooperative game theoretical techniques for Energy Aware Task scheduling

Abstract

This paper proposes energy-efficient scheduling algorithms for distributed heterogeneous grids. The conservation of energy consumption has a tirade of effects that improve system reliability, increases in the lifespan of the system, ideally with only a linear degradation in performance. With a few restrictions and proper modeling, we convert the Energy Aware Task Allocation (EATA) problem into a Bargaining game, wherein the machines compete with each other. Although we have done work previously in addressing the EATA problem, the question arises whether the players should cooperate or not. In this paper, we compare cooperative and non-cooperative games for the EATA problem on a heterogeneous set of machines. The contribution of the paper is not just the comparison of the two approaches but also two new algorithms for the EATA problem with energy consumption as the primary objective. The energy savings generated by the proposed algorithms is achieved through Dynamic Voltage and Frequency Scaling (DVFS) at the processor level. The comparison leads to some interesting results, highlighting the strength of each scheme.