Resilient and Low Latency Communications in Smart Grid Environments

The decarbonization of the power supplies, with the introduction of renewable energies, new kinds of loads, such as smart vehicles, are making the electric power system a dynamic environment. The ancient assets, the legacy software and protocols used for the control of the power grid are not enough anymore, and the introduction of new ICT technologies is bringing the concept of smart grid, allowing the orchestration and the integration of the efforts of the main actors of the power system (generators, carriers, and consumers). This is translated in the need of increasing the observability of the power grid, which can be achieved via a huge network of sensors, interacting among them and with the infrastructure in order to provide information about the physical world. This data can be later stored, processed, analysed in order to control the behaviour of the grid through intelligent actuators or for offline analysis. The smart grid networks should manage a great amount of data, delivered over different physical media, coming from many different types of devices, some of them with limited computational power, and with different requirements in terms of QoS. This thesis work analyses the usage of a publish-subscribe model in the context of a smart grid environment, underlining the aspects of data availability, durability and reliability, keeping latencies under control. Moreover, the design of the ICT of a distribution system as a data-centric architecture is provided, a place where data can be asynchronously produced, consumed, processed, aggregated, and stored, without the need of direct interactions between producers and consumers.