MVDR broadband beamforming using polynomial matrix techniques

Weiss, S. and Bendoukha, S. and Alzin, A. and Coutts, F.K. and Proudler, I.K. and Chambers, J.; (2015) MVDR broadband beamforming using polynomial matrix techniques. In: European Signal Processing Conference 2015. IEEE, FRA, pp. 839-843. ISBN 978-0-9928626-3-3 (https://doi.org/10.1109/EUSIPCO.2015.7362501)

[thumbnail of Weiss-etal-EUSIPCO-2015-MVDR-broadband-beamforming-using-polynomial-matrix]
Preview
Text. Filename: Weiss_etal_EUSIPCO_2015_MVDR_broadband_beamforming_using_polynomial_matrix.pdf
Accepted Author Manuscript

Download (417kB)| Preview

Abstract

This paper presents initial progress on formulating minimum variance distortionless response (MVDR) broadband beamforming using a generalised sidelobe canceller (GSC) in the context of polynomial matrix techniques. The quiescent vector is defined as a broadband steering vector, and we propose a blocking matrix design obtained by paraunitary matrix completion. The polynomial approach decouples the spatial and temporal orders of the filters in the blocking matrix, and decouples the adaptive filter order from the construction of the blocking matrix. For off-broadside constraints the polynomial approach is simple, and more accurate and considerably less costly than a standard time domain broadband GSC.

ORCID iDs

Weiss, S. ORCID logoORCID: https://orcid.org/0000-0002-3486-7206, Bendoukha, S., Alzin, A. ORCID logoORCID: https://orcid.org/0000-0001-8716-8233, Coutts, F.K., Proudler, I.K. and Chambers, J.;