A RANK-(<I>L</I>, <I>L</I>, 1) BCD BASED AOA-POLARIZATION JOINT ESTIMATION ALGORITHM FOR ELECTROMAGNETIC VECTOR SENSOR ARRAY

This paper investigates an angle of arrival (AOA) and polarization joint estimation algorithm for an L-shaped electromagnetic vector sensor array based on rank-(L,L, 1) block component decomposition (BCD) tensor modeling. The proposed algorithm can take full advantage of the multidimensional information of electromagnetic signal to obtain the parameter estimation more accurately than the matrix-based method and the existing tensor decomposition method. In addition, the algorithm can accomplish pair-matching of estimated parameters automatically. The numerical experiments demonstrate that even under the conditions of low SNR and limited snapshots, the proposed algorithm can still steadily achieve high detection probability with low estimation error, which is important for practical applications.