A Deep Learning Method for NLOS Error Mitigation in Coastal Scenes

With the widespread use of automatic identification systems (AISs), some ships deceptive information or intentionally close their AISs to conceal illegal activities evade supervision maritime departments. Although radar measurements can be effectively utilized evaluate credibility received AIS data, propagation non-line-of-sight (NLOS) signal conditions is an important factor that affects location accuracy. This study addresses NLOS problem in a special geometric dilution precision (GDOP) scenario on coast and several base stations. We employed data augmentation deep residual shrinkage network order alleviate adverse effects errors. The results our simulations demonstrate proposed method outperforms other range-based localization algorithms mixed LOS/NLOS environment. For GDOP with four radars, algorithm’s root-mean-square error (RMSE) was lower than 180 m.