The Modified Model of Unknown Correlated Noise Fields under the Polarization Sensitive Array

Abstract

Currently for the polarization vector array of unknown noise covariance matrix still use scalar matrix model, resulting in low precision of source number estimation Estimation and a series of problems. This paper defines the polarization sensitive array under unknown correlated noise field of the noise covariance matrix, it is derived under the assumption that the spatial noise covariance is block diagonal or banded, covariance matrix of the original noise model is modified. Three models are compared based on scalar matrix, polarization sensitive vector array, modified polarization array vector array underlying unknown correlated noise field. The simulation results of MATLAB show that the correct detection probability of the modified model can be significantly improved. Furthermore the analysis of noise suppression is envisaged due to the fact that modified noise covariance matrix has Toeplitz structure. 