Control Barrier Function Based Decentralized UAV Swarm Navigation While Preserving Connectivity Without Explicit Communication

This paper proposes a decentralized leader-follower navigation method for a group of unmanned aerial vehicles (UAVs) in environments with obstacles. The proposed method is based on control barrier functions (CBFs) and is designed to overcome the limitations of current approaches that use artificial potential functions (APFs), which may in some cases cause undesired vibratory movements. The algorithm presented in this study computes control inputs using only local sensor information and without explicit communication between the robots, with the aim of ensuring collision avoidance, obstacle avoidance, and connectivity preservation in the group. The control inputs are derived without relying on numerical solution techniques, and the proposed approach aims to prevent constraint violations from occurring before they happen. The performance of the proposed approach is compared to the APF-based approach in the context of vibratory movements, and simulation results demonstrate its ability to achieve smoother robot movements. Moreover, the effectiveness of the proposed approach is confirmed through experimental validation with quadrotors.