Computational Microwave Imaging With a 1-bit Phase Reconfigurable Transmitarray

Recently, computational microwave imaging (CMI) based on randomized measurements has been validated effective to achieve low-cost, high-resolution imaging. In this work, we present a CMI system with a phase-reconfigurable transmitarray (PRTA). The main contributions are threefold. First, a low-cost PRTA capable of achieving 1-bit, 0/
$\pi$
phase modulation is realized for optimal imaging with a minimal insertion loss of 0.5 dB. Second, a discrete-dipole approximation model of the PRTA is derived to predict the fields radiated onto the imaging plane. By doing so, the transfer matrix can be quickly established by numerical calculations, eliminating the need of time-consuming calibration measurements, and enabling real-time and flexible experimental imaging with desired imaging distance and resolution. Third, a randomly focused field illumination strategy was used for CMI in this work, reducing the interference reflections from surrounding objects and improving the signal-to-noise ratio (SNR) of measured signals. An experimental prototype working at 24 GHz with a 16
$\times$
16-element PRTA was realized, showing imaging results that capable of distinguishing two objects separated by 4 mm (about 0.32
$\lambda)$
, indicating a sub-wavelength scale imaging resolution.