Title :
Estimation of the number of sources based on the eigenvectors of the covariance matrix
Author :
Nezafat, Mahdi ; Kaveh, Mostafu ; Xu, Wenyuan
Author_Institution :
Dept. of Electr. Eng., Minnesota Univ., Minneapolis, MN, USA
Abstract :
In this paper, we propose a method for estimating the number of sources based on the eigenvectors of the covariance matrix of a narrowband array. The proposed method provides the same detection sensitivity as AIC and MDL, but appears to be much more robust to signal and noise model perturbations. Unlike AIC and MDL that do not make use of the procedure used in calibrating the antenna array, the proposed method can use the calibrated antenna responses to take into account the existing nonidealities present in the antenna array, and hence outperforming AIC and MDL methods in practice.
Keywords :
antenna arrays; array signal processing; calibration; covariance matrices; eigenvalues and eigenfunctions; signal detection; signal sources; AIC; Akaike information criterion; MDL; antenna array calibration; covariance matrix; detection sensitivity; eigenvector; minimum description length; narrowband array; source estimation; Antenna arrays; Covariance matrix; Direction of arrival estimation; Directive antennas; Eigenvalues and eigenfunctions; Multiple signal classification; Narrowband; Noise robustness; Receiving antennas; Signal processing;
Conference_Titel :
Sensor Array and Multichannel Signal Processing Workshop Proceedings, 2004
Print_ISBN :
0-7803-8545-4
DOI :
10.1109/SAM.2004.1502991