Underwater acoustic positioning using reconstructed wavefront sets based on time-of-arrival in multipath environments

Abstract Inspection of underwater structures in ports requires accurate positioning; however, existing methods suffer from severe multipath in very shallow waters. Existing CFAR-based approaches depend on precise tuning of the threshold, which is highly sensitive to environmental conditions. This study proposes a robust underwater acoustic positioning method that integrates multiple peaks detected by several hydrophones and identifies the direct path by evaluating the spatial scatter of back-propagated wavefronts. By using sphere intersections based on estimated distances, the method provides more reliable direct-path identification than CFAR-based techniques. The proposed approach was validated through image-source simulations and sea trials conducted in Tateyama, Japan. The results show that it reduces mean positioning errors and significantly decreases the maximum error compared with existing methods. These findings demonstrate that the proposed method is robust to multipath and environmental variations, offering a practical, easy-to-use solution for automated underwater inspection of port structures.