Augmented correlation matrix reconstruction for direction of arrival estimation using dual-spread array

Due to the large degrees of freedom (DOF) by generating difference coarray, nonuniform arrays have great potential in direction of arrival (DOA) estimation. However, some existing nonuniform arrays (e.g., (sparse) nested arrays) still have comparatively strong mutual coupling in physical array domain, whereas the others (e.g., coprime arrays) have holes or only used part available in difference coarray domain, for the performance loss of DOA estimation. To address these issues, we propose two kinds of dual-spread arrays which achieve more and continuous DOF by exploiting correlation matrix reconstruction, rather than the generation of difference coarray. All inter-antenna spacings of the proposed arrays are larger than half-wavelength, resulting in extended aperture and reduced mutual coupling. An ESPRIT-based estimation method is further provided for unique DOA estimation with high-accuracy and low-complexity. The analyses of mutual coupling and computational complexity are provided in detailed. The theoretical analyses and simulation results show the superiority of the proposed arrays and method over the existing techniques.