Direction-Finding Algorithm for Non-Circular Signal by a Two-Level Large-Spacing Symmetric Nested Array

To mitigate mutual coupling and enhance the degrees of freedom (DOFs), this paper proposes a two-level large-spacing symmetric nested array (TLSSNA). Unlike conventional nested arrays, the proposed configuration does not restrict the minimum inter-element spacing to the half-wavelength of the incident signal. In conjunction with the array geometry, a direction of arrival (DOA) estimation algorithm tailored for non-circular signals is developed. Under the adopted mutual coupling model, the proposed large-spacing geometry reduces the performance degradation of the proposed algorithm caused by mutual coupling. The proposed algorithm first constructs multiple correlation matrices utilizing both difference and sum co-arrays, which are subsequently integrated into a partitioned correlation matrix. Eigenvalue decomposition (EVD) is then performed on the combined matrix to derive an initial signal subspace. This subspace is further partitioned, extended, and recombined to yield a new subspace equivalent to the signal subspace of a uniform linear array (ULA). Ultimately, high-precision and unambiguous DOA estimates are obtained via the estimation of signal parameters via rotational invariance techniques (ESPRIT). Simulation results demonstrate that the proposed algorithm achieves superior estimation accuracy over conventional nested arrays and other existing large-spacing array geometries.