Enhanced Long Baseline Underwater Target Localization with Adaptive Track-Before-Detect Method

In recent years, the particle filter (PF)-based track-before-detect (TBD) method has garnered attention in long baseline (LBL) localization algorithms. This approach can overcome the measurement-to-track association (MTA) challenges and complex underwater environments. In this paper, LBL underwater target localization capability is enhanced by designing the likelihood ratio function and constructing adaptive thresholds. Specifically, our contributions are as follows: First, we design the likelihood ratio function to enable automatic tracking management decisions and reduce the convergence time. Second, we construct adaptive thresholds to cope with the dynamically changing environment. Based on simulation results, the proposed algorithm outperforms the traditional localization algorithm and PF-TBD algorithm in dynamically changing environments, especially when signal-to-noise ratios are low, and has superior tracking and location capabilities.