Bayesian modeling of acoustic signals for seafloor identification

In this paper the Helmholtz-Kirchhoff approximation for backscattering strength and a Bayesian model of the uncertainty related to acoustic backscatter measurements are integrated into a maximum a posteriori processing scheme for the estimation of seafloor roughness parameters. Two processors are developed based on different levels of uncertainty regarding the angles of incidence for the received acoustic signals and the system calibration. Simulations indicate that the new maximum a posteriori processors are superior to a maximum likelihood estimation scheme that operates under the assumption that the angles of incidence for the backscattered signals are fixed and treats the calibration factor in a deterministic fashion. Specifically, the new processors produce roughness parameter estimates which are very close to the true values of the parameters, which are known in simulations, whereas the fixed angle processors are shown to result in a substantial bias in the estimation procedure.