Subspace-based single-vector-sensor direction finding for two coherent acoustic sources having real-valued constellations

Perfect correlations between the three particle velocity components and acoustic pressure of an acoustic source effect an efficient aperture for a solitary acoustic vector-sensor which is able to simultaneously collect the acoustic pressure and three orthogonal particle velocity components. Single-vector sensor direction finding for non-coherent signals is thus possible. For coherent signals, source decorrelation is required by many computationally attractive direction finding algorithms such as spatial filtering methods and subspace-based methods to recover the rank property of the source covariance matrix. Conventional source decorrelation techniques such as spatial smoothing and angular smoothing necessitate averaging of several shifted invariance subarrays, which is impossible for one acoustic vector-sensor. To solve the problem of rank deficiency, a novel method called phasic smoothing is herein suggested and validated for source decorrelation and direction finding of two coherent noncircular signals with one acoustic vector-sensor.