Super-Resolution Joint Azimuth-Range-Doppler Estimation via OFDM Waveforms

Journal article, 2026

This article considers the problem of super-resolution estimation of multidimensional target parameters using orthogonal frequency division multiplexing (OFDM) waveforms. Existing super-resolution techniques primarily focus on either angle estimation or range-Doppler estimation. Attempting to obtain high-dimensional parameter estimates by separately estimating angle, range, and Doppler and then pairing the results often leads to parameter mismatches and target identification ambiguity. To address this issue, we propose an extended propagator method that employs a 3-D affine subarray mapping to fully exploit intersource phase variations and avoid rank deficiency in multidimensional problems. By vectorizing the received data and applying forward-backward smoothing, the cross-dimensional estimation problem is reformulated as a multivariate phase-rotation estimation problem. The final parameters are obtained via an orthogonal-propagator-based spatial spectrum search, achieving accurate high-resolution estimation without covariance matrix construction or eigenvalue computation. To further reduce computational complexity, a dimension-reduced sequential estimation framework is introduced. Under the considered simulation settings (e.g., standard 5G configurations with limited data size and moderate-to-high SNR), the proposed approach reduces computational complexity by approximately two to three orders of magnitude, at the cost of only about a 10% increase in RMSE. The proposed method is theoretically formulated in a general form, and a two-stage implementation is illustrated using Doppler estimation following range-angle estimation. Both simulations and experiments based on a 5G-band OFDM integrated sensing and communication platform validate the effectiveness of the proposed approach.