SDN-Based Management of Wide Area Measurement System (WAMS) and Traffic Optimization Mechanism for Smart Grid Communication

Abstract

Electricity has become one of the most fundamental resources of human life. The Supervisory control and data acquisition (SCADA) system was initially used in smart grid monitoring and is still present in traditional and semi-automated GRIDS. RTUs, meters and protective relays send the system measurements to the SCADA framework. The SCADA framework intermittently surveys estimating phasor data. SCADA takes 2-10 seconds to perform calculations of phase angle data. In addition to that, the SCADA system cannot assign timestamps to phase angle data. The Wide area measurement system is introduced in the Smart Grid monitoring System. The Wide Area Measurement system (WAMS) is more accurate and fast as compared to the SCADA framework. The WAMS architecture is specifically designed for monitoring purposes. The Phasor Measurement Unit (PMU) is attached to GPS (Global Positioning System), so the time and location of the PMU are fully accurate and synchronized. PMUs are efficient devices which can calculate the synchro phasor data accurately along with the timestamp. PMU can send 120 frames per second. When the Phasor Data Concentrator (PDC) receives PMU frames, it aligns them based on timestamps and sends them to the control center for processing. Our purposed SDN based communication model can be deployed within and among the substations and lead towards the regional and main control centers. In Smart Grid infrastructure, substations are the ending points of HAN and NAN and starting points of WAN. In Smart Grid, WAN is equipped with Wide Area Measurement devices such as PMUs, PDS and Communication Network. In our proposed SDN based model, SDN gateway switches are maintained at substations. However, these switches are controlled by SDN local controllers, local controllers placed to reduce the load from global controllers and network nodes. These local controllers are fully synchronized with the global controllers in order to maintain the updated network state. A Queuing priority mechanism is also proposed for important and delay-sensitive data. Simulation results show that the network load is optimized, and obtained jitter is less than 10ms by the proposed model.