Skip to Main Content
The polarimetric angular smoothing (PAS) technique for coherent direction-of-arrival (DOA) estimation and adaptive beamforming is explored and enriched by taking into account the cross-correlations among the smoothed subarrays and using non-uniform weights. The PAS using annihilating weights is shown to be free of signal cancellation [which occurs in the conventional spatial smoothing (SS) technique for uncorrelated sources] while effectively suppressing coloured noise with unknown Toeplitz covariance matrix, but may be incompetent for DOA estimation. The PAS technique is still shown to be more attractive than the SS scheme in a very difficult situation that the direct signals and the reflecting signals are of very close DOAs. The penalised Toeplitz-block-Toeplitz scheme is further suggested for more thorough source decorrelation with an L-shaped electromagnetic vector-sensor array to gain improved interference suppression. The analytical expression for the mean-squared error of DOA estimates obtained from the weight-PAS pre-processed root-multiple signal classification is derived and analysed. A hybrid DOA estimator incorporating PAS and integrating some existing techniques is also described. Numerical results are finally given.