INS/AOA Fusion for Dual-UAV Relative Positioning: An Asymptotically Efficient Closed-Form Solution

Relative positioning is critical for collaborative operations among unmanned aerial vehicles (UAVs). This study proposes a relative positioning algorithm for Dual-UAV that utilizes inertial navigation systems (INS) and opportunistic communication signals to obtain angle-of-arrival (AOA) measurements in Global Navigation Satellite System (GNSS)-denied environments. Both AOA measurements and INS-derived attitude and displacement measurements are corrupted by noise, introducing statistical bias relative to the Cramér-Rao lower bound (CRLB). To mitigate this bias, we calibrate weighting matrix parameters using relative position estimates, enhancing accuracy through iterative refinement. Theoretical analysis and simulations demonstrate the algorithm asymptotically attains CRLB performance.