Extended physical layer modeling for smart metering utility network simulators

Abstract: Smart grids are nowadays considered to be the most promising electricity distribution framework,leading to a greener and more cost-eective electricity consumption than before. Their technologicalforefront consists in advanced communication systems allowing full-duplex communications till the veryedge of distribution networks, which also turn smart grids into platforms for a new variety of homeautomation services. One key innovation in such a communication system consists in how meters accessthe smart grid data network. Within this context, the Smart metering Utility Networks (SUNs) standardhas been recently developed. As an emerging technology allowing to deploy smart grid distributionsystems, SUNs are considered a relevant but yet largely unexplored research topic. Within this work, a simulation tool for assessing the performance of SUNs is developed and employedto study a typical scenario. In order to obtain a high coherence with real networks in terms of theirsize and the reproduced OSI layers, a network simulator (ns-3) has been chosen as main development environment. The focus is on the physical layer, which is typically represented by a coarse structure in network simulators, lacking most of the relevant communication features. Since network simulators typically do not allow to fully implement physical layer features as in scientic programming languages(e.g. Matlab), we develop models reflecting physical layer behavior onto the above layers. The main engaged challenges regarding model construction are an adequate modeling of data units structure, channel properties and transmission features, including advanced transmission techniques such as forward error correction coding and multiple-antenna beamforming. Simulation runs prove our tool to be reasonably valid and consistent with the analyzed technology. Furthermore, they provide a basis for formulating design guidelines for reliable communication schemes inthe considered scenario. The developed models constituting the tool show a general structure that can be employed for any wireless network simulator and, due to their proven validity, open paths to interestinginvestigation and optimization work.