Numerical performance of a tensor music algorithm based on HOSVD for a mixture of polarized sources

Mélanie Boizard (1) , Guillaume Ginolhac (2) , Frédéric Pascal (3) , Sébastian Miron (4) , Philippe Forster (1)

1 SATIE - Systèmes et Applications des Technologies de l'Information et de l'Energie
2 LISTIC - Laboratoire d'Informatique, Systèmes, Traitement de l'Information et de la Connaissance
3 SONDRA - Sondra, CentraleSupélec, Université Paris-Saclay (COmUE)
4 CRAN - Centre de Recherche en Automatique de Nancy

In this paper, we develop an improved tensor MUSIC algorithm adapted to multidimensional data by means of multilinear algebra tools. This approach allows to preserve the multidimensional structure as the signal and the noise subspaces are estimated from the Higher Order Singular Value Decomposition (HOSVD) of the covariance tensor. The proposed algorithm is applied to a polarized source model. By computing the Mean Squared Error (MSE) for different scenarios, the performance of this method is compared to the classical MUSIC algorithm as well as the vector MUSIC algorithm that includes the polarization information. The simulations show that our algorithm outperforms the vector algorithms.

Mots clés: Tensor MUSIC; HOSVD; DOA Polarimetric sources estimation

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

Mélanie Boizard, Guillaume Ginolhac, Frédéric Pascal, Sébastian Miron, Philippe Forster. Numerical performance of a tensor music algorithm based on HOSVD for a mixture of polarized sources. 21st European Signal Processing Conference (EUSIPCO 2013), Sep 2013, Marrakech, Morocco. pp.CDROM. ⟨hal-00871220⟩