Matched-field minimum variance beamforming in a random ocean channel
Matched-field source localization methods that employ deterministic full- wave acoustic propagation models can be seriously degraded due to the presence of random inhomogeneities in the ocean channel. In this paper, a minimum variance (MV) matched-field beamformer is presented that achieves greater robustness to random inhomogeneities in the sound-speed profile between the source and receiver. The proposed modification of the MV beamformer consists of employing multiple linear constraints derived from predicted pressure fields obtained using a set of perturbed sound-speed profiles. In order to investigate the nature of wave-front variations due to random sound-speed perturbations, a normal mode model based on adiabatic and first-order perturbation approximations is examined. The signal wave-front spatial correlation implied by this model suggests that the coherence among modes can remain high even in a fluctuating ocean environment. This in turn implies that the dimension of the signal perturbation constraint space for the MV beamformer can be small for typical sound-speed variations at moderate source ranges. Given the signal constraint space, design of the MV beamformer with sound-speed perturbation constraints is achieved by selecting its quiescent response to maximize the average signal-to-noise ratio gain against spatially uncorrelated noise. This leads to a computationally efficient realization of the beamformer that avoids the need to repeatedly compute perturbed pressure fields. Simulation experiments using a realistic deep- water Pacific Ocean environment are presented, which suggest that robust unambiguous low-frequency source location estimates can be achieved in the presence of mesoscale inhomogeneities given only knowledge of the second- order statistics of the random range-dependent sound-speed profile plus a single environmental measurement at the receiving array.
Volume / Issue
Start / End Page
International Standard Serial Number (ISSN)
Digital Object Identifier (DOI)