Modeling of reverberation from the seabed for Doppler sonar

In the last 30 years, low-deviation Doppler sonar became a critical tool for autonomous high-precision underwater navigation. This technology estimates the velocity of a platform by extracting Doppler information from reverberation. This study proposes a method for modeling reverberation by constructing a seabed with random undulations to investigate the formation mechanism of measurement errors, optimize sonar design, enhance accuracy, and quantify estimation quality. Initially, the statistical characteristics of measured reverberation and velocity are analyzed to establish a more realistic scattering process and comprehensively simulate the velocity uncertainty in a random channel. Then, a reverberation model is developed based on the K-distribution exhibited by the measured reverberation, simulating the spatial pattern of Gaussian speckle with specified density and scale. Finally, the effectiveness of the model is verified through theoretical and experimental means. The simulated velocity exhibits significant channel dependence and aligns with the uncertainty observed in the measured velocity, which goes below the Cramer-Rao lower bound. The model effectively simulates the reverberation and the measured velocity distribution.