Off-grid sparse iterative adaptive approach for joint range-angle estimation in FDA-MIMO radar

Automotive radar has become an indispensable piece of advanced driver assistance systems, yet resolving densely distributed targets in urban traffic scenarios remains a critical challenge. In such scenarios, conventional radar systems suffer from degraded perception reliability under stringent real-time constraints. To address this challenge, we investigate the use of frequency diverse array (FDA) multiple-input-multiple-output radar for high-resolution joint range-angle estimation, and a efficient two-stage, coarse-to-fine estimation strategy tailored for practical off-grid conditions is proposed. In the first stage, the time-frequency characteristics of linear frequency-modulated waveform are leveraged to mitigate FDA range ambiguity and extracts single-snapshot signals from range bins. Subsequently, within the vicinity of these coarsely estimated bins, a refined estimation is obtained using the proposed off-grid sparse iterative adaptive approach (OGSIAA). OGSIAA addresses the off-grid model mismatch by approximating the true dictionary matrix via a first-order Taylor expansion. Furthermore, it incorporates an adaptive weighted sparse regularization term, optimized by the Bayesian information criterion to promote robust sparse recovery while maintaining computational efficiency. Extensive numerical simulations confirm that the proposed framework achieves superior estimation accuracy and resolution in off-grid cases with low computational complexity.