Evaluating Trail Formation in Collaborative UAV Networks with Lethal Threats

Abstract

Aerial vehicles are widely used for missions in hostile environments. Some of these missions can benefit from the use of unmanned aerial vehicles (UAVs) working collaboratively to achieve a common goal. In such scenarios, the use of stealth approaches is essential for mission success, because it allows missions to be conducted in the proximity of threat radars and also reduces the probability of a UAV being spotted by enemies by enemies. In this direction, we extend a model that allows autonomous UAVs to dynamically and collaboratively reduce their exposure to threats by adopting a stealth policy in which they turn off their communication radars once a threat is detected. The model also encompasses strategies for UAVs leaving the threat regions as quickly as possible and estimating the future position of others UAVs, thus trying to regroup to formation. The model was evaluated in simulations using different hostility and weapon range distributions. The results demonstrated that the proposed model is able to significantly improve the number of UAVs that are able to complete the mission without being hit by any threat. In addition, the proposed approach for calculating the goal position when a UAV is exposed to a threat proves to be a key factor for missions performed in hostile environments as its significantly reduces the fraction of UAVs hit by threats.