Passive TOA Localization for Maritime Surveillance Using AIS Signals

Maritime localization traditionally relies on cooperative systems such as the Automatic Identification System (AIS), where vessels actively transmit their positions. However, AIS spoofing (i.e., the intentional falsification of transmitted coordinates) undermines the reliability of these approaches. This paper investigates Time of Arrival (TOA)-based passive localization as an independent, physical-layer solution for vessel position verification. A key challenge in this scenario is the presence of an unknown timing offset that is common to all receiving base stations (BSs) for each AIS transmission, resulting from transmitter delays, protocol timing references, and receiver processing latencies. To address this issue, the localization problem is formulated by explicitly incorporating the unknown temporal bias into the measurement model. Several estimation strategies (either iterative or not) are then investigated/designed, including Maximum Likelihood (ML), Least-Squares (LS), and constrained formulations, to solve the resulting nonlinear problem. Experimental validation using real AIS datasets shows that the considered approaches enable accurate passive vessel localization using existing AIS infrastructures. In particular, all the considered methods provide a favorable trade-off between estimation accuracy and computational complexity, while satisfying requirements for real-time applicability.