Analysis of TCP Traffic in Smart Grid Using SDN Based QoS

Smart Grid (SG) communication infrastructure is expected to be scalable, efficient and adaptable to support inevitable grid growth. Software Defined Network (SDN) is a technology that has so far proven it usage in scalable distributed systems especially for traffic prioritization management and is thus a prime candidate for the SG network traffic management that is capable of reaching 100% utilization. In this paper, SG communication network is emulated on a physical testbed and the performance of the Automatic Generation Control (AGC) Smart Grid application performance under high network load when using traditional and SDN networks is compared. The AGC network traffic performance is analyzed w.r.t. duplicate acknowledgement, packet retransmission, packet loss and congestion windows size. Congestion window size was evaluated for four algorithms: Cubic, Westwood, Veno and Scalable. The results show that SDN based QoS management significantly improves overall performance of the critical SG application.