Sparse Bayesian learning-based spatial spectrum estimation for mobile sonar platforms during turning

Abstract

Mobile sonar platforms usually experience turning in the course of travel. During turning, due to the change of heading, the steering vector of the sonar array is not constant, which will cause the widening of the spatial-spectral peaks or even fail to estimate target bearings. To deal with this problem, this paper first establishes a multi-snapshot fusion equation for observed data from different heading angles. Then, a sparse Bayesian learning-based method is utilized to solve the fusion equation and provides the estimate of the spatial spectrum. The simulation results exhibit that the proposed method can resolve the port and starboard ambiguity problem and provide high estimation accuracy with enough heading change information. The sea trial results validate its feasibility and stability for estimating far-field target bearings in practical applications.