Energy Management for Cloud Computing: A survey from Scheduling Perspective of Heuristic, Game Theory and Learning Strategy

Cloud computing has been a champion in Information Technology (IT) world for supporting flexibility on-demand services. It is leveraged to instantly fulfill dynamic demands from heterogeneous environment. However, the massive demand for the Cloud computing services contributed to high energy consumption, thus affecting cost of service and processing components lifetime. Quite a portion of the energy consumed is dissipated for ensuring processing components (i.e., processor) running but idle. An effective task scheduling is the solution for better processing performance while reducing energy consumption. In this paper we compared several energy efficient scheduling from three (3) different perspectives (i.e., heuristic, economic and intelligent points of view). These different perspectives are chosen particularly for exemplifying essential principles and features of energy management in task scheduling. Novel knowledge on the energy management are identified through similarities, differences and energy models. We believed that better trade-off between energy consumption and system performance is the solution to green Cloud computing.