A Distributed Multi-UAV System for Search and Tracking of Unauthorized UAV Operators via RF–Vision Fusion

Unauthorized uncrewed aerial vehicle (UAV) operations in low-altitude airspace require localization of the ground-side operator, not only detection of the airborne platform. This article presents a distributed multi-UAV system that searches for and tracks an unknown number of unauthorized UAV operators using radio frequency (RF) and visual sensing. The system combines decentralized probability hypothesis density estimation, posterior fusion, and adaptive coordination under clutter, missed detections, and communication constraints. RF sensing provides long-range passive cues from control-link emissions, whereas vision provides accurate short-range observations. Each UAV fuses the two posterior intensities locally and exchanges only compact posterior summaries with its neighbors, avoiding the transmission of raw sensor data. The fused intensity guides decentralized motion planning toward informative regions and improves time difference of arrival (TDOA) sensing geometry. Simulations across different target densities, team sizes, and sensing ranges show lower optimal subpattern assignment (OSPA) error than single-modality and distributed-control baselines.