Single-chip near-field MIMO-SAR imaging system: development and experimental validation

Abstract

Millimeter-wave near-field imaging offers high resolution, non-ionizing radiation, and partial penetration through some low-loss nonmetallic coverings, making it highly valuable for concealed metallic target detection. However, conventional near-field SAR imaging relies heavily on dense mechanical sampling, resulting in relatively low acquisition efficiency. In addition, when operating in the extreme near field, single-chip MIMO radars face challenges such as inter-channel amplitude and phase inconsistencies, static background interference, and multistatic geometric mismatch, all of which can degrade imaging quality. To address these issues, this paper develops a single-chip near-field MIMO-SAR closed-loop imaging system for concealed target detection and establishes a processing pipeline integrating channel calibration, background subtraction, multistatic-to-monostatic phase correction, and frequency-domain reconstruction. The proposed system enables closed-loop coordination of radar configuration, two-dimensional mechanical scanning, raw echo acquisition, and data management, thereby ensuring a stable correspondence between spatial sampling positions and multichannel echo data. Experimental results demonstrate stable focusing and effective complex-target reconstruction under sparse mechanical sampling, while preserving imaging capability for concealed metallic targets across diverse geometries and occlusion conditions, including cardboard, clothing, and double-layer clothing–cardboard coverings.