Acoustic Source Localisation

Acoustic Source Localisation

Performance of certain direction of arrival (DOA) estimation algorithms using a uniform linear array of scalar sensors has been analyzed and reported. Velocity hydropones, which measure one Cartesian component of the three-dimensional particle velocity vector of the incident wave field, were first reported in underwater acoustics literature in 1956. But the importance of simultanously measuring both pressure and particle velocity was recognized nearly three decades later when Mann etc. published a classic paper addressing the fundamentals of energy transfer in acoustic field.

The relevance of this new approach was further elucidated by GL D'Spain. The technology of vector hydrophones or acoustic vector sensors (AVS), consisting of three orthogonally oriented velocity hydropones plus a pressure hydrophone, all spatially collocated, for simultaneous measurement of pressure and particle velocities is of relatively recent vintage. It has attracted considerable interest, especially within the signal processing community, after Nehorai and Paldi presented the AVS measurement model and a method for localizing acoustic sources using an array of such sensors. Hochwald and Nehorai later addressed the important issue of identifiability with vector sensors, namely, the bound on the number of sources identifiable in a class of array processing models with multiple parameters and signals per source. Hawkes and Nehorai have quantified the impact of sensor locations on the performance (Cramer Rao Bound) of an AVS array.

In a separate paper they have defined some performance measures such as mean square angular error (MSAE) and mean square range error (MSRE) and derived asymptotic lower limits on these quantities in terms of the CRB. The same authors have also discussed the correlations between the measurements within an AVS as well as between two spatially separated acoustic vector sensors. They have also analyzed the effect of a reflecting boundary and presented a fast method for localizing a wideband source using a distributed AVS array.

A host of beamforming techniques using vector sensors have also appeared in the literature. Wong and Zoltowski introduced ESPRIT-based closed form source localization for arbitrarily spaceed 3D arrays of vector hydropones. They have also proposed aperture extension using a uniform rectangular array of vector hydrophies spaced much farther than half a wavelength to achieve enhanced array resolution and direction-finding precision. Narrowband Beamforming and Capon direction estimation methods with an AVS array have been discussed by Hawkes and Nehorai. Root-MUSIC based azimuth-elevation DOA estimation technique using uniformly spaced velocity hydropones has been developed by Wong and Zoltowski. They have also proposed a blind MUSIC-based source localization algorithm using arbitrary 3D array of vector hydropones. The same authors have also presented a near-field / far-field azimuth-elevation angle estimation algorithm using a single vector hydrophone.

Leave a Reply

Your email address will not be published.