Extended-Aperture ESPRIT-Based 2-D direction finding with a sparse uniform array of electromagnetic vector sensors

Aperture extension is achieved in this novel ESPRIT-based two-dimensional (2-1)1 angle estimation scheme using a sparse uniform rectangular array of electromagnetic vector-sensors spaced much farther apart than a half wavelength. An electromagnetic vector-sensor is composed of six spatially co-located, orthogonally oriented, diversely polarized antennas, distinctly measuring all six electromagnetic-fleld components of an incident multisource wavefleld. Each incident source's direction of arrival is determined from the source's electromagnetic-fleld vector components, which are extracted by decoupling the data correlation matrix's signal-subspace eigenvectors using the lower-dimensional eigenvectors obtainable by TLS-ESPRIT. These direction-cosine estimates are unambiguous but have high variance; they are used as coarse references to disambiguate the cyclic phase ambiguities in ESPRIT's eigenvalues when the intervector sensor spacing exceeds a half wavelength. In one simulation scenario, the estimation standard deviation decreases linearly with increasing intervector sensor spacing up to 16 wavelengths, effecting a 97% reduction in the estimation standard deviation. In another simulation scenario, the proposed scheme and attendant electromagnetic vectorsensor array outperform a uniformly spaced array of unpolarized antennas with the same aperture and a slightly greater number of component antennas, lowering the signal-to-noise ratio (SNR) resolution threshold by 9 dB and achieving a 21-dB SNR performance gain. Other simulations further demonstrate the significant performance gain realizable by this method for wireless mobile fading-channel communications.