Title :
SINR loss of the dominant mode rejection beamformer
Author :
Wage, Kathleen E. ; Buck, John R.
Author_Institution :
Electr. & Comput. Eng. Dept., George Mason Univ., Fairfax, VA, USA
Abstract :
The signal to interference and noise ratio (SINR) characterizes a beamformer´s ability to attenuate unwanted signals. SINR loss quantifies how close the performance of a beamformer designed with sample statistics is to the performance of a beamformer designed with ensemble statistics. This paper conjectures that the SINR loss of the dominant mode rejection (DMR) beamformer is beta-distributed and provides substantial numerical evidence supporting this claim. Building on results from random matrix theory, the paper proposes a simple perturbative model for the sample covariance eigenvectors, which are known to control DMR performance [1]. SINR loss for the proposed model agrees with the beta distribution conjecture.
Keywords :
array signal processing; covariance matrices; eigenvalues and eigenfunctions; DMR beamformer; SINR loss; beta distribution conjecture; dominant mode rejection beamformer; ensemble statistics; random matrix theory; sample covariance eigenvectors; signal to interference and noise ratio; simple perturbative model; unwanted signal attenuation; Array signal processing; Covariance matrices; Histograms; Interference; Monte Carlo methods; Signal to noise ratio; Standards; adaptive arrays; beamforming; random matrix theory; sample covariance matrix;
Conference_Titel :
Acoustics, Speech and Signal Processing (ICASSP), 2015 IEEE International Conference on
Conference_Location :
South Brisbane, QLD
DOI :
10.1109/ICASSP.2015.7178421