Angle of arrival estimation of plane waves propagating in random media

Abstract

The well-studied problem of estimating the angle of arrival of plane waves in ocean acoustics is revisited. While beamforming is the classical technique, the most recent methods use parametric models of the signals, and maximum likelihood estimation (MLE). Yet, in all cases, fluctuations in the required signals due to medium perturbations have been neglected. In this paper, the advantages of using the correct second moment for waves propagating in random media (WPRM) to model the fluctuations of the arriving signals are investigated. In this way, the physics of WPRM is incorporated in this parametric model of the observations. This model is then used as the basis for the MLE. Finally, the Cramer–Rao lower bound versus fluctuation strength, angle of arrival, and the noise power is provided. Also developed are iterative ML algorithms based on the Newton–Raphson method for a single source and extended to multiple sources. The method is applicable for plane wave arrivals at any known array geometry with arbitrary medium autocorrelation function. Simulation results show a dramatic improvement in the precision of the estimated angle of arrival.