Particle-Velocity Coarray Augmentation for Direction Finding with Acoustic Vector Sensors

In this paper, the problem of passive direction finding is addressed using an acoustic vector sensor array (AVS), which may be deployed either in free space or near a reflecting boundary. Building upon the 4x1 vector field measured by an AVS, the particle-velocity coarray augmentation (PVCA) is proposed to admit the underdetermined direction finding using the spatial difference coarray derived from the vectorization of the array covariance matrix. Unlike the widely used spatial coarray Toeplitz recovery technique, the PVCA is applicable to arbitrary array geometries and imposes no reduction of the spatial difference coarray aperture. For the array located at or near a reflecting boundary, the PVCA allows resolving up to 13 sources, while for the array located in free space, the PVCA can identify 9 sources at most. By applying to the systematically designed nonuniform arrays, such as coprime arrays and nested arrays, the PVCA can be coupled with the spatial smoothing technique to get the number of resolvable sources multiplied. Finally, the effectiveness of the PVCA is verified by numerical simulations.

Automated summary

In this paper, the problem of passive direction finding using an AVS is addressed. The PVCA method is proposed to overcome the underdetermined direction finding using the spatial difference coarray. Unlike existing techniques, PVCA can be applied to arbitrary array geometries and does not reduce the spatial difference coarray aperture. It is shown that PVCA can resolve up to 13 sources for arrays near a reflecting boundary and 9 sources for arrays in free space.