Self-initiating velocity-field beamspace MUSIC for underwater acoustic direction-finding with irregularly spaced vector-hydrophones

This paper introduces a novel MUSIC-based (MUltiple Signal Classification) blind source localization algorithm applicable to three-dimensional arbitrarily spaced arrays of velocity-hydrophone triads. This proposed algorithm (1) self-generates coarse estimates of the sources' arrival angles to start off its MUSIC-based iterative search without any a priori source parametric information, (2) exploits information embedded in the impinging sonar velocity-field (as versus pressure field), (3) automatically pairs the x-axis direction-cosine estimates with the y-axis direction-cosine estimates. This method uses vector-hydrophones, each of which comprises three spatially co-located but orthogonally oriented velocity-hydrophones. Each velocity-hydrophone distinctly measures one Cartesian component the incident sonar wavefield's velocity-vector. Velocity-hydrophone technology is well established in underwater acoustics and a great variety of commercial models have long been available. This proposed algorithm forms velocity-field beams at each vector-hydrophone, and uses coarse estimates of each source's velocity-vector estimate obtained by decoupling the signal-subspace eigenvectors. Simulation results verify this innovative scheme's capability to self-generate initial direction-cosine estimates for its MUSIC-based iterative search and demonstrate the proposed algorithm's superior performance relative to a similarly spaced army of pressure-hydrophones. Under one scenario, the proposed method lowers the estimation bias by 95% and the estimation standard deviation by 47%, relative to a similarly configured array of pressure-hydrophones provided with a priori initial arrival angle estimates.