On secure UAV collision avoidance

Kurzfassung

Over the last decade, connected vehicle research has extended from ground transportation into the airspace, with drones anticipated to play a central role in shaping the urban airspace ecosystem. Larger number of flights however also increase the risk of collisions proportionally. While some threats can be mitigated through Unmanned aeronautical vehicle Traffic Managment methods, such as conflict-free flight plans, an independent last line of defense is essential. DroneCast aims to address this need by providing a collaborative collision avoidance system based on broadcasted position reports of all drones. However, to prevent malicious actors from disrupting orderly traffic flow by injecting ghost targets that cause unnecessary avoidance maneuvers, robust cybersecurity measures are required. This paper recaps strategies for securing DroneCast's broadcast messages given its resource-constrained environment. Beyond establishing robust cybersecurity measures, it further explores approaches for handling messages which can not be verified at the time and highlights their feasibility to deploy in real-world environments, laying the groundwork for future research.