Deep-Ocean Transient Source Localization Using Arrival Time Differences Matching and Adaptive Weighted Discrete Fourier Transform

We propose a transient source localization method using arrival time differences (ATDs) matching and adaptive weighted discrete Fourier transform (AWDFT) in the deep ocean with only one hydrophone. For a single hydrophone placed in the shadow zone, the common practice of the transient acoustic source localization is extracting and matching ATDs of the received data with calculated ATDs of the copy field. ATDs can be acquired from the interference structure in frequency domain, which is formed because of overlapping multipath signals, by implementing discrete Fourier transform (DFT). However, traditional DFT can't give satisfactory ATD estimations because of using a fixed window function with a low resolution and high sidelobes, which will degrade the localization performance. Considering that, we propose an improved DFT method, called AWDFT, to improve the ATD estimation performance. AWDFT can choose optimal window functions adaptively from a group of predesigned windows, thereby provide a higher resolution and lower sidelobes. Numerical simulations validate AWDFT can hold a more precise ATD estimation and a better noise suppression ability simultaneously comparing to traditional DFT, and with the help of AWDFT, the proposed localization method can give localization results with satisfactory accuracy and robustness.