A low-complexity and energy-efficient dynamical adjustment strategy for servers in a data center

In order to meet the exponential exploding traffic growth trends, several hybrid optical-wireless access network architectures are proposed to provide better services and gain healthy profit. Data center is an important component in many network architectures. To provide high quality of service (QoS) while avoiding great energy consumption trade-off at the same time, researching on energy saving for data center is becoming more and more essential. However, conventional methods usually ignore the complexity, causing many of them cannot be put into practice at the present stage. In this paper, we focus on both energy efficiency and the complexity of servers' adjustment in a data center by proposing an effective and easily realized strategy. In addition, the energy consumption distribution of the data center under our strategy is given, and Monte Carlo method is employed to verify it. The simulation results show that our proposed strategy consumes only about 5% more energy while reduces 98.7% of calculation complexity compared to the theoretical optimal energy-efficient strategy, thus justifying its further application.