Research On The Data Transmission Delay In The Smart Grid Communication Networks

Smart Grid（SG）communication networks are the fundamental support for therealization of future Smart Grid. In the Smart Grid communication networks, there aredifferent kinds of data which include hard real-time data, quasi-real time data as wellas non-real time data. As for the requirements of transmission delays, they can rangefrom a few milliseconds to several seconds. Therefore, it has been a great problem tosolve the transmission delay for different kinds of data in the Smart Gridcommunication networks, and it is also a core fundamental to ensure the Smart Gridcommunication networks’ security, reliability, efficiency as well as intelligence.In our paper, based on the background of Smart Grid, firstly we build astatistical model for the data in the Smart Grid communication networks, and analyzethe data characteristics for different kinds of data. In the following, on the basis of therelated researches and works, we formulate the latency of transmission and breakdown its components. By mathematical analysis, we find that the waiting time forqueuing is the key factor for the problem.In order to solve this problem, in our first step, we propose the VL-EDF（Virtual Latency Earliest Deadline First） scheduling algorithm by introducing theconcept of “Virtual Latency”. By analysis and simulation, we find that our queuescheduling algorithm can better support the latency-sensitive data in the Smart Gridcommunication networks. And compared with the other scheduling algorithms, ourproposed VL-EDF algorithm can satisfy the latency requirements for all kinds of dataunder heavier network traffic loads.Base on the VL-EDF queue scheduling algorithm, a latency based flowscheduling algorithm LBFS is further proposed. In the algorithm, all of the switchnode （such as switches, routers） ports are monitored. Once the network traffic rateof a certain port is beyond the predefined threshold, we will compute new routes, withthe minimum latency cost, for the latency-insensitive traffic flows so that the heavylatency-insensitive traffic flows are moved to the light-load routes. By analysis andsimulation, the results show that LBFS can better solve the latency caused by thecongestion, and we also find that LBFS algorithm can better improve the latencyperformance of traffic flows, especially for the latency-sensitive traffic flows. Atmeanwhile, it has little impact on the affected flows in the new links.At last, we briefly introduce the Software Defined Networking （SDN）, and develop the LBFS scheduling algorithm on the SDN platforms. With the experimentsand results, it shows that LBFS scheduling algorithm can improve the latencyperformance of the traffic flows in the Smart Grid communication networks. LBFSdecreases the latency that caused by the congestion, and better satisfies thelatency-sensitive data’s latency requirements.