UAV collision avoidance exploitation for noncooperative trajectory modification

Distributed collision-free trajectories are generally obtained through a continuous sharing of information between vehicles. With the intent of investigating possible sources of vulnerability in autonomous frameworks, we formalize a procedure malicious players can follow to influence other. In this paper we propose a strategy for steering a UAV towards predetermined targets. The strategy described here relies on the existence of a flight information sharing protocol (i.e. ADS-B) and predictable collision avoidance algorithms. A model predictive controller is applied to the switching system representing a pair of UAVs coupled by the presence of an imminent collision. As showed by means of numerical simulations and robot experiments, the result is a loss of autonomy on the UAV. Our results suggest the need to include the subject of our study in the discussion on safe automated airspace.