Wireless Networks for Smart Grid Communications: Traffic Perspective and Resource Allocation

Abstract

The Smart Grid (SG) is the next-generation electrical power grid. It aims to provide reliable and efficient energy distribution and consumption by utilising information and communications technologies. At the core of the SG is the integration of communications networks with the power grid. However, the bottleneck issue in these networks is the last mile connection, which means connecting the substations and customer premises to the backhaul communications network via the so-called last mile access networks, referred to as SG access networks. Due to the wide coverage of these networks, spanning both the distribution and transmission levels of the power grid, they will incorporate numerous SG devices, generating a vast amount of SG communications traffic with various requirements. In this thesis, we assess different deployments of the SG access networks by using accurate analytical frameworks, based on queuing systems. Specifically, we develop analytical frameworks for three possible deployments considered by the energy utilities, referred to as public, private, and hybrid SG access networks. In addition, an analytical framework for a deployment of SG access networks, which is based on the concept of wireless network virtualisation, is developed. The accuracy of the derived analytical expressions is validated by simulations of a wireless network model with realistic SG traffic profiles obtained from the Australian utility operator (Ausgrid). The analytical and simulation results are shown to agree very well. Thus, these frameworks might be applied by the energy utilities and telecommunications operators as tools for designing SG access networks that satisfy their communications requirements. Finally, to meet the ever-increasing demands on the radio spectrum, arising from numerous SG devices and various applications, a novel scheduling and resource allocation algorithm is proposed for wireless SG networks.