BOTTOM REVERBERATION SUPPRESSION USING DELAY-DOPPLER DECONVOLUTION FOR THE DETECTION OF LOW-SPEED SMALL TARGETS IN VERY SHALLOW WATER

In the scenarios of detecting the underwater low-speed small targets (e.g., frogmen or UUVs) in the very shallow water environment (e.g., harbor or shoal where the water depth is about several meters), much attention has been attracted on reverberation suppression. Based on the Doppler difference between the moving target echoes and the bottom reverberation dominant in the very shallow water, the traditional matched filtering may realize the separation of the bottom reverberation and the target echoes. However, the Doppler incurred by low-speed small targets is quite small, so the performance on the bottom reverberation suppression seriously degrades due to the limitation of Doppler resolution of the traditional approach. Therefore, it is urgent to improve its Doppler resolution to separate low-speed small targets out from the nearly static bottom reverberation. Generally, the reverberation and the target echoes can be considered as the weighted sum of the copies of transmitted signal with different delay and Doppler, which indicates that “the cross-ambiguity function of the received echo is approximately the two-dimensional convolution of the auto-ambiguity function of transmitted signal and the generalized “target” reflectivity density function (i.e., the weighted sum of the 2-D Dirac Delta function convolved with different delays and Doppler shifts)”. In this paper, regarding the auto-ambiguity function of the transmitted signal as the 2-D point spread function (PSF), an iterative deconvolution algorithm called the Richardson-Lucy (R-L) algorithm is exploited to deblur the conventional delay-Doppler output. This processing can obtain approximately the generalized “target” reflectivity density function similar to the thumbtacks, which can significantly improve the resolution of the delay and Doppler of the raw output, and separate the slow-speed small targets submerged by the bottom reverberation. Simulations show the effectiveness and superiority of the proposed reverberation suppression method.