Direction of arrival estimation for conformal arrays with diverse polarizations

Polarization diversity of conformal antenna arrays introduces uncertainty to the determination of array manifold needed for passive direction finding. Most of the previous research on direction of arrival (DOA) estimation for arrays with diversely polarized elements focused on the algorithms to estimate both the directions and polarizations of the incident waves[1][2][3]. In many applications however, knowledge of polarizations is not of interest and therefore it's possible to eliminate the uncertainty and complexity incurred by the diversity in polarizations and beam patterns among array's elements. In this article two data models are proposed for conformal array's DOA estimations. One is for single incident signal scenario, by treating the physical conformal array as one consisting of virtual uniform radiators through post-processing of received data. Another is a polarization-sensitive array manifold for DOA estimation of multi-incident signals. Simulation results validate the proposed algorithm, which has improved robustness and significantly lower computational requirements in comparison with traditional data models. The proposed algorithm is of great value in practical applications, especially those requiring real-time processing.