Gridless DOA Estimation for Coprime Array in GNSS Spoofing Environments

Direction-of-arrival (DOA)-based antispoofing techniques provide significant security enhancements for global navigation satellite system (GNSS) receivers by mitigating spoofing threats. This paper proposes a coprime array-based gridless DOA estimation framework tailored for GNSS spoofing environments. The denoised covariance matrix is first vectorized prior to receiver despreading. This transformation enables the construction of a difference coarray containing spatial gaps, which is subsequently interpolated to restore continuity. We then apply atomic norm minimization for obtaining the underlying signals from the interpolated virtual array. We synthesize a Toeplitz matrix extracting a non-zero row from rank-deficient matrix. By applying the estimation of signal parameters via rotational invariance techniques (ESPRIT) into a reconstructed Toeplitz matrix, this method enables direct derivation of DOA estimates in closed-form, circumventing the need for computationally intensive off-grid error compensation. Numerical simulations validate the approach, demonstrating its superiority in maximizing degrees-of-freedom utilization while achieving enhanced estimation accuracy compared to conventional grid-based techniques.