Model order estimation of highly correlated signals applied on measured data

Abstract

The paper deals with the model order estimation of a covariance matrix in the case of two coherent incoming signals on an antenna array. The paper also deals with the estimation of the angle of arrival (AoA). The used model order estimation techniques, are Minimum descriptive length (MDL), information theoretic criteria (AIC) and method which uses pseudo-spectrum of the direction of arrival (DoA). In the paper a motivation of the model order estimation in the AoA tasks is explained and the problems state of the art is described. Several methods of model order estimation are presented, their analysis executed and an experimental verification of the selected methods is performed.