Joint Azimuth–Elevation Super-Resolution Estimation via 2-D Iterative Weighted Atomic Norm Minimization

The rapid development of unmanned aerial vehicle (UAV) swarm technology, characterized by small size, low radar cross-section and coordinated motion dynamics, poses significant challenges to conventional radar systems in target detection and high-precision localization. Effective detection and localization of swarm targets are hindered by the limited resolution and low echo signal-to-noise ratio (SNR). To address these challenges, this paper proposes a joint azimuth–elevation super-resolution estimation method based on two-dimensional iterative weighted atomic norm minimization (2D-IWANM). A unified processing framework is established by integrating long-time coherent integration (LTI) with sparse reconstruction techniques, providing an effective solution for radar-based UAV swarm localization. Based on this framework, the LTI technique enhances the echo SNR and compensates for the signal phase, establishing a reliable basis for subsequent processing. The 2D-IWANM algorithm operates directly on the original signal matrix, avoiding the computational cost of vectorization in conventional 2D vectorized atomic norm minimization (2D-VANM) methods. Furthermore, an adaptive iterative weighting strategy is incorporated to update atomic weights in each iteration, effectively suppressing spurious peaks, overcoming resolution limitations, and enhancing the ability to resolve closely spaced targets. Simulation and comparative results show that the proposed algorithm achieves superior resolution, robustness and computational efficiency. Its effectiveness is further validated using two sets of measured data.