AN OPTIMAL PRIOR KNOWLEDGE-BASED DOA ESTIMATION METHOD

Guillaume Bouleux (1) , Petre Stoica (2) , Remy Boyer (3)

1 LASPI - Laboratoire d'Analyse des Signaux et des Processus Industriels
2 Uppsala University
3 L2S - Laboratoire des signaux et systèmes

In some applications involving line spectra or direction of arrival (DOA) estimation, we may have a priori information which could consist of known frequencies in Magnetoencephalography (MEG) or mechanical signals, or of known DOA's in a RADAR urban sce- nario. With this fact in mind we propose an optimal method for array processing that exploits the information of the known DOA's for es- timating the unknown DOA's as accurately as possible. This new Prior-knowLEDGE (PLEDGE) technique is based on the method of direction estimation (MODE) approach. To show the benefits of incorporating prior-knowledge we also present the corresponding stochastic Crame ́r-Rao Bound (CRB). Finally, we use PLEDGE for estimating frequencies in the current spectrum of an induction motor to perform the diagnosis on rotor bars.

Mots clés: DOA estimation; MODE; prior knowledge; maximum likelihood; fault diagnosis; frequency estimation; Cramer-Rao bound

Domaines: Traitement du signal et de l'image [eess.SP]; Traitement du signal et de l'image [eess.SP]

Guillaume Bouleux, Petre Stoica, Remy Boyer. AN OPTIMAL PRIOR KNOWLEDGE-BASED DOA ESTIMATION METHOD. 17th Europeen Signal Processing Conference, EUSIPCO, Aug 2009, Glasgow, United Kingdom. pp.117. ⟨hal-00456900⟩