A Network Traffic Scheduling Strategy for Energy Storage Data Centers Based on SDN

With the advancement of the times, the combination of energy storage power stations and data centers (DCs) has become increasingly close. Software-defined network (SDN), as an advanced network architecture, is also gradually applied to the data center network (DCN), which plays a huge role in the network traffic scheduling. DC has the specific topology and traffic characteristics. Therefore, in the context of energy storage DC, we propose a network traffic scheduling strategy for processing long flows based on SDN in this paper. In the strategy, long flows can be divided into two parts: aggregated long flows and non-aggregated long flows. We use different algorithms to schedule these two kinds of flows. The aggregated long flows here refer to the long flows with the same source and destination. We compare our proposed strategy with the other classical traffic scheduling algorithms, the simulation results perform better in average throughput, load balancing, and link bandwidth utilization.