An Efficient Two-Stage Near-Field 3-D Imaging Algorithm for Massive MIMO Radar With Range Migration and ADMM

The near-field imaging technique for massive multiple-input multiple-output (MIMO) radar has received increasing attention in applications such as personnel security screening, nondestructive evaluation, and medical imaging due to its potential for high-resolution reconstruction. However, existing techniques face significant challenges in simultaneously achieving high computational efficiency, superior image quality, and low memory consumption. These trade-offs pose difficulties for practical system deployment. To address the aforementioned trade-offs, this paper proposes a two-stage near-field imaging algorithm for massive MIMO arrays. In the first stage, the equivalent phase center principle is employed to transform MIMO data into single-input single-output (SISO) data associated with virtual elements. A fast reconstruction is then performed using the range migration algorithm (RMA), although the resulting image quality remains suboptimal. Furthermore, both the forward and inverse imaging operators of this stage are analyzed in detail. In the second stage, the imaging operators are utilized to model the imaging process, and the image is reconstructed via the alternating direction method of multipliers (ADMM) to avoid computationally intensive large-scale matrix operations. Additionally, a composite image-quality metric is introduced to terminate the iterative process. Simulation and experimental results prove the effectiveness of the proposed algorithm.