Research On Energy Saving Strategy Of Campus Data Center Based On SDN

In recent years,with the vigorous development of the Internet of Things and mobile networks and the ever-increasing demand for Internet services,a large number of data centers have emerged.However,the huge energy consumption during data center work leads to the increase of operating costs and carbon emissions.Therefore,the problem about energy conservation of data centers has become one of the focuses of research.In this paper,three cardinal issues about traffic forecasting,network monitoring and energy saving strategies are specific researched based on the SDN architecture.Aiming at the problem of energy waste during low data flow in campus data center,a dynamic and adaptive energy saving strategy is proposed based on improved traffic forecasting algorithm,and in order to verify effectiveness of the improved algorithm,a new network monitoring model is designed ulteriorly in this paper.Firstly,a traffic forecasting algorithm is improved based on the gray theory using the Markov transition matrix through the analysis of the historical data of a university data center in Zhejiang Province.The experimental results show that the absolute value of the mean daily relative error of the improved algorithm is 3.32%which is less than the error threshold,so this method has certain reference value for the reasonable deployment of data center in the future.Secondly,in order to verify the validity of the forecasting algorithm,a network monitoring model in which Mininet and RYU controller used to build data center network resource pool is designed to realize real time monitoring for the traffic conditions and host state.The experimental results show that the improved algorithm in this paper is more accurate than the traditional engineering flow forecasting algorithm through analyzing the data which collected by the model,so the improved algorithm can allocate network resources more rationally and optimally to achieve the effect of energy saving.Finally,a dynamic and adaptive energy saving strategy which combined forecasting algorithm and the monitoring model is proposed to solve the problem of energy waste during the low data flow in campus data center.This strategy which takes into account the actual work schedule and traffic load of the campus uses exclusive routing and particle swarm optimization routing respectively according to two different traffic situation of data center in winter or summer vacations and school opening to achieve the optimization of the number of devices(components)and improve the performance of data center energy-saving.