Around-the-corner Radar Sensing Using Reconfigurable Intelligent Surface

Around-the-corner radar (ACR) sensing of targets in non-line-of-sight (NLOS) conditions has been explored for security and surveillance applications and look-ahead warning systems in automotive scenarios. Here, the targets are detected around corners without direct line-of-sight (LOS) propagation by exploiting multipath bounces from the walls. However, the overall detection metrics are weak due to the low strength of the multipath signals. Our study presents the application of reconfigurable intelligent surface (RIS) to improve radar sensing in ACR scenarios by directing incident beams on the RIS into NLOS regions. Experimental results at 5.5 GHz demonstrate that micro-Doppler signatures of the walking motion of humans can now be captured in NLOS conditions through the strategic deployment of RIS.

💡 Research Summary

The paper investigates the use of a reconfigurable intelligent surface (RIS) to enhance around‑the‑corner radar (ACR) sensing in non‑line‑of‑sight (NLOS) environments. Traditional ACR relies on multipath reflections from walls, which are inherently weak and lead to ambiguous localization because multiple propagation paths can satisfy the same geometric constraints. To overcome these limitations, the authors design and fabricate a 1‑bit coding metasurface RIS consisting of a 16 × 10 array of unit cells (overall size 256 mm × 160 mm, unit‑cell periodicity 16 mm). Each cell uses a PIN diode to provide a 180° ± 20° phase shift between its ON and OFF states, yielding a fractional bandwidth of 18.5 % centered at 5.45 GHz.

The experimental platform employs a FieldFox N9926A vector network analyzer (VNA) configured as a monostatic radar. The VNA transmits a continuous‑wave signal at 5.5 GHz with a modest 10 kHz bandwidth and +3 dBm power, while sampling at the maximum 370 Hz to avoid Doppler aliasing. Each measurement lasts 3 seconds, producing roughly 4000 complex samples that are later processed with a short‑time Fourier transform (STFT) using a 0.3 s Hamming window to generate micro‑Doppler spectrograms.

The test environment is an L‑shaped indoor corridor (width 2.4 m, brick walls 30 cm thick). The radar and RIS are mounted at a height of 1.1 m; the RIS is placed 1.8 m from the radar with an incidence angle of –60° relative to the negative x‑axis, allowing the reflected beam to be steered to 30°, 45°, or 60° with respect to the normal. Four human walking trajectories are examined: (T1) tangential LOS, (T2) tangential NLOS, (T3) radial LOS, and (T4) radial NLOS.

Baseline measurements without RIS show strong torso returns and weak arm/leg micro‑Doppler in LOS cases (T1, T3), while NLOS cases (T2, T4) produce barely discernible spectrograms due to severe path loss and multipath smearing. When the RIS is introduced, NLOS spectrograms become clearly visible. At 30° and 45° steering angles, a human walking 2 m from the RIS yields detectable Doppler bursts between 1 s and 3 s, despite the two‑way propagation (radar → RIS → target → RIS → radar) which incurs higher attenuation than one‑way communication links. The 60° steering angle shows reduced performance because the reflected beam does not fully illuminate the target region. Distance dependence is also evident: moving the target from 2 m to 3 m away from the RIS reduces reflected power noticeably, highlighting the limited field‑of‑view (≈11°–12° 3‑dB beamwidth) of the current RIS design.

The authors conclude that RIS can substantially improve detection metrics in ACR by providing a directed, low‑loss path into NLOS zones, thereby mitigating the weakness of pure multipath. However, the experimental setup also reveals constraints: the narrow 10 kHz radar bandwidth limits range resolution; the 1‑bit RIS offers limited beamforming flexibility and a modest aperture, restricting coverage; and the two‑way reflection doubles propagation loss. Future work is suggested to (i) develop multi‑bit or continuous‑phase RIS for finer beam steering, (ii) integrate wideband radar waveforms to exploit range‑Doppler diversity, and (iii) expand the RIS aperture or employ multiple RIS panels to enlarge the effective field of view, especially for multi‑target or longer‑range scenarios. Overall, the study demonstrates a practical, low‑cost pathway to bring RIS technology into radar‑based security, automotive look‑ahead warning, and smart‑city sensing applications.