Title :
Blind focusing wide-band array processing
Author :
Bendjama, A. ; Bourennane, S.
Author_Institution :
S.D.E.M., Univ. of Corsica, Corte, France
Abstract :
The coherent signal subspace method based on the second-order statistics has been proposed for estimating the directions of arrival (DOA) of multiple wide-band sources. These techniques are ineffective in presence of the Gaussian noise and also when the propagation model is unknown. In this paper, the extension of the focusing algorithms to the noisy data is developed. This extension is based on the higher-order statistics cumulant. Indeed, the use of the fourth order cumulant leads to eliminate the Gaussian noise contribution and then, to improve the high resolution algorithms on the one hand, on the other hand to determine the unbiased focusing matrices needed for estimating the coherent signal sub-space. To improve the cumulant matrix estimation, a spatial average of slice cumulant matrices is developed. The performances of our algorithm are studied by the use simulated data.
Keywords :
Gaussian noise; array signal processing; direction-of-arrival estimation; matrix algebra; statistical analysis; DOA estimation; Gaussian noise; coherent signal subspace method; cumulant matrix estimation; direction-of-arrival estimation; higher-order statistics cumulant; second-order statistics; wide-band array processing; Standards;
Conference_Titel :
Signal Processing Conference, 2000 10th European
Conference_Location :
Tampere
Print_ISBN :
978-952-1504-43-3