MUSIC and Improved MUSIC Algorithms for Parameter Estimation Using a Polarization Sensitive Array

Abstract

This paper considers the multiple signal classification (MUSIC) and improved MUSIC algorithms for parameter estimation using a polarized antenna array. The signal direction of arrival (DOA) and polarization parameters are estimated using these algorithms. Reduced dimension methods are proposed to lower the computational complexity. These methods are applied to a uniform linear array and a two-dimensional uniform L-array, and both coherent and incoherent incident signals are considered. Results are presented which show that the MUSIC algorithm can accurately estimate the parameters of incoherent signals with both arrays. The improved MUSIC algorithm can estimate all parameters with the L-array but some cannot be obtained with the linear array. For coherent signals, the MUSIC algorithm fails to estimate the DOA. The improved MUSIC algorithm can accurately estimate the DOA with the linear array, but the error is large with the L-array. Further, it cannot provide the polarization parameters of coherent signals with either array. The reason is that the autocorrelation matrix of the received signals is no longer Toeplitz, so it cannot be reconstructed using the properties of Toeplitz matrices.