Radar QR Code Based on Metasurfaces With Phase-Gradient Dispersion

Coded radar reflectors are increasingly employed in applications requiring reliable identification, tracking, and improved signal integrity. In contrast to conventional active transponders, passive solutions offer greater robustness by placing the task of information retrieval on the radar interrogation system, thereby supporting the joint communication paradigm of enhancing the capabilities of existing channels. We introduce passive radar QR codes composed of engineered metasurface pixels, designed to interface with automotive FMCW radar systems widely deployed on modern vehicle platforms to enhance road safety. Designed for the 76-81 GHz automotive range, the metasurface pixels generate a strong reflection phase gradient across the band. We demonstrate that FMCW signal processing, which maps phase information to estimated range, interprets a phase chirp as an artificial shift in the detected target distance. Consequently, metasurfaces with distinct phase chirps encode information analogous to Braille, enabling radar systems with adequate angular resolution to decode them as QR codes. We validated this concept by designing and fabricating several phase-chirped metasurfaces on PCB substrates and characterizing their performance in an anechoic chamber. The single-ring metasurface design achieves approximately 330° of phase excursion and yields a measured artificial range shift of 7.43 cm over a 500 MHz band centered at 80 GHz. Multiple QR code patterns were constructed and successfully detected. To expand the dataset, data augmentation techniques were applied, and a custom-designed convolutional neural network (CNN) demonstrated strong detection capabilities for the proposed system. On an augmented dataset of 1024 QR code samples represented as 224×224×3 heatmaps, a ResNet50V2-based CNN exceeds 85% classification accuracy at SNR values of 10-15 dB and achieves a pixel position mean absolute error below 10% at SNR values above 15 dB. Radar QR codes hold promises for serving as passive Vehicle Identification Numbers (VINs), enabling direct interrogation by standard automotive radar systems and paving the way for secure, infrastructure-independent identification in future autonomous mobility scenarios.