5G Edge Offloading for Drone Swarms

Over the past few years, we have experienced a significant evolution in cloud technologies and paradigms, such as the "Edge Computing" model, which allows to deliver quick responses by offering services at the network's edge, thus significantly reducing the latency in favor of responsiveness. In order to find a solution that could efficiently leverage the computational capability present at the network's edge, the “Flexible, scaLable, secUre, and decentralIseD Operating System” (FLUIDOS) European project has been proposed, which has as its main goal the creation of a seamless computing continuum that can be used to integrate heterogeneous devices, thus creating a single virtual cluster between the IoT devices, such as drones, and the network’s edge. This work aims at enhancing FLUIDOS capabilities by providing the appropriate support for ensuring seamless operations even in fluctuating network conditions, like in 5G networks, through the “Distributed Edge Analytics Service” (DEAS) open call. The thesis will discuss the design and implementation of the solution required to achieve an autonomous, adaptive, and efficient workload distribution that allows to dynamically offload tasks between drones and the edge by using the quality of the 5G channel as a decision criterion, thus ensuring a responsive computing continuum. As a conclusion, the thesis will evaluate the performance of the proposed solution and will discuss the possible future directions of the project.