Impact of packet loss on power demand estimation and power supply cost in smart grid

The evolving smart grid will use advanced data communications and networking techniques to improve efficiency and reliability of electric power generation, transmission, distribution, and consumption. Packet loss performance of the data communications networks used in smart grid will have impact on the cost of power supply. In this paper, we model and analyze the impact of packet loss performance on the optimization of cost for power supply in smart grid. To optimize (i.e., minimize) the cost of power supply, the power demand from consumers has to be estimated based on the power usage data, which are transferred from smart meters through data aggregator unit (DAU) to the meter data management system (MDMS). First, we present a model to optimize the cost of power supply given demand uncertainty. Then the probability distribution of power demand is estimated with and without packet loss. Subsequently, we analyze and show how the packet loss increases the cost of power supply. Next, a queueing model is used to quantify the packet loss due to congestion at DAU, and then the transmission rate from the DAU is optimized to minimize the impact of packet loss. The modeling and analysis of packet loss performance presented in this paper is a step toward optimal network design for future smart grid.