Successive Method for Angle-Polarization Estimation With Vector-Sensor Array

A successive method for joint angle and polarization estimation of multiple narrowband incoherent sources using linear electromagnetic vector-sensor (EMVS) arrays is proposed in this article. First, the polarization field smoothing root-multiple signal classification (MUSIC) algorithm is presented to estimate the elevation angles. Second, the estimated elevation angles are applied to extract the EMVS manifolds and then to estimate the azimuth angles. Finally, the estimated angles are used to provide the polarization parameter estimates. Compared with most of the previous algorithms, the proposed algorithm first uses the polarization field vector smoothing, which enables the offering of more accurate elevation angle estimates for correlated sources, with no loss of array aperture, and, second, requires no spectral searching and parameter pairing procedures. Simulations results are provided to show the efficacy of the proposed algorithm.

I. INTRODUCTION

Joint angle and polarization estimation (JAPE) of multiple diversely polarized signals is a common problem encountered in radar, communications, and seismic sensing. For example, in radar applications, accurate angle and polarization information extraction offers better target localization, classification and recognition performance. During the past decades, many efficient JAPE methods using electromagnetic vector sensor (EMVS) arrays have been developed [1]– [13]. Among these methods, the eigenstructure-based methods, such as the Multiple Signal Classification (MUSIC)-based algorithms [7], [8] and the Estimation of Signal Parameters Via Rotational Invariance Techniques (ESPRIT)-based algorithms [4]– [6], have become popular due to their high estimation performance.