Enhanced sonar bathymetry tracking in multi-path environment

In this paper we address DOA estimation for the side scan sonar in the presence of multiple interfering echoes. We illustrate the potential usage of high resolution methods and tracking algorithms. The proposed tracking algorithm is based on a apriori information on the sea-floor DOA angle. Because of the non-linearity of the model and non-Gaussian behavior of the observed 4600 data, the implementation of the proposed algorithm is based on the particle filter. The proposed tracking algorithm is shown to be able to resolve the multi-path interference problem. The heavy-tailed/non-Gaussian character of the data is noted and the Laplace distribution is shown to better characterize the tails of the observed data. The multivariate Laplace distribution is derived for the observed data and the particle filter coupled with the multivariate Laplace distribution is shown to provide better estimates than with the Gaussian assumption.