Large-array acoustic signal processing in random deep-water channels: Effects of coherence

Abstract

Approximate analytical solution of the matrix transport equation for the mutual coherence function (MCF) of an acoustic field in terms of normal modes was derived. The proposed approach is capable of including stochastically rough surface as well as volume fluctuations in the index of refraction, and allowed one to reduce the problem of the MCF calculation in a refractive channel to the analogous problem in a free space. These results have been then employed to simulate the statistical effects of long-range sound propagation in realistic deep-water environments from the North-West Pacific on the horizontal array detection performance for several beamforming techniques. The most attention was paid to the gain degradation for the sound frequency of 250 Hz over megameter ranges under the basic assumption that internal waves or surface wind waves are, respectively, the main source of acoustic signal fluctuations. Rough surface scattering has been shown to cause the most significant effects. In particular, for the array length of order 100 acoustic wavelengths and wind speed of 15 m/s the coherence loss was of about 6 dB even for optimal quadratic processor and, in comparison, of about 12 dB for conventional plane-wave beamformer. [Work supported by Russian Basic Research Foundation, Grant No. 94-02-04544a.]