3-D Near-Field Source Localization Using a Spatially Spread Acoustic Vector Sensor

Abstract

This article presents a new method for 3-D localization of an acoustic source signal by estimating its azimuth-angle, elevation-angle, and radial range. The proposed method exploits a spatially spread acoustic vector sensor, which is composed of a tri-axial velocity vector sensor and an isotropic pressure sensor. Unlike the spatially collocated case, the self-normalization of spatially spread acoustic vector sensor’s response is no longer independent of the source location, resulting in the inapplicability of the widely used “self-normalization” estimators (A. Nehorai <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">et al.</i> , 1994), (V. N. Hari <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">et al.</i> , 2012), (Y. Song <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">et al.</i> , 2013). The present article considers the near-field propagation’s path attenuation and phase difference among the sensor components and derives a source location estimation method without resorting to the “self-normalization” operation. Also, the proposed method is applicable to any nonfree-space propagation models at arbitrarily unknown path-loss exponent. In comparison with the existing methods, where the velocity vector sensor is spatially collocated, the proposed method can offer improved performance estimation due to the inherent extension of the vector sensor’s spatial aperture in the spread structure.