⚡ electrical engineering

Around-the-corner Radar Sensing Using Reconfigurable Intelligent Surface

This paper demonstrates that deploying a Reconfigurable Intelligent Surface (RIS) significantly enhances around-the-corner radar sensing in non-line-of-sight conditions by strengthening multipath signals, enabling the successful capture of human micro-Doppler signatures at 5.5 GHz.

Original authors: Kainat Yasmeen, Debidas Kundu, Shobha Sundar Ram

Published 2026-02-13

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Kainat Yasmeen, Debidas Kundu, Shobha Sundar Ram

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The "Magic Mirror" That Lets Radar See Around Corners

Imagine you are standing in a hallway with a flashlight, trying to spot a friend who is walking around a corner. You can't see them because the wall blocks your light. This is the classic problem of "Non-Line-of-Sight" (NLOS) detection.

For decades, scientists have tried to solve this with radar. Instead of light, they use invisible radio waves. The old way of doing this is like shouting into a canyon and listening for an echo. The radio waves bounce off the walls, travel around the corner, hit your friend, bounce back to the wall, and finally return to the radar.

The Problem: This "echo" is very weak. By the time the signal makes all those bounces, it's like trying to hear a whisper in a hurricane. The radar gets confused, the image is fuzzy, and it's hard to tell if the person is walking, running, or standing still.

The Solution: Enter the "Smart Mirror" (RIS)

This paper introduces a new gadget called a Reconfigurable Intelligent Surface (RIS). Think of the RIS not as a passive wall, but as a high-tech, programmable mirror.

The Old Way: The radar waves hit a rough brick wall and scatter in every direction, like a ball bouncing off a crumpled piece of paper. Most of the energy is lost.
The New Way: The RIS is a flat panel covered in tiny, switchable tiles (like a grid of microscopic mirrors). When the radar beam hits it, the RIS can "catch" the signal and focus it like a laser pointer, aiming it precisely around the corner toward the hidden person. Then, it catches the return signal and focuses it straight back to the radar.

It's the difference between shouting into a crowd and using a megaphone to speak directly to one person.

How They Tested It

The researchers set up a real-world experiment in a building with an L-shaped corridor (a corner).

The Setup: They placed a radar on one side of the corner and a human "target" (a person walking) on the other side, completely hidden from the radar's direct view.
The Test:
- Without the Mirror: They tried to detect the person using only the weak, scattered echoes from the walls. The result? The radar could barely see anything. The "signature" of the person's walking (the way their arms and legs moved) was lost in the noise.
- With the Mirror: They placed the RIS panel strategically. Suddenly, the radar could "see" the person clearly.

The "Micro-Doppler" Secret Sauce

How did they know it was a person and not just a random noise? They looked at the Micro-Doppler signature.

Imagine a spinning fan. Even if the fan is moving slowly, the blades are moving fast. Radar can detect this tiny, rapid movement. Similarly, when a human walks, their torso moves slowly, but their arms and legs swing back and forth quickly. This creates a unique "fingerprint" in the radar signal.

Without RIS: The signal was too weak to see the swinging arms. It was just a blurry blob.
With RIS: The signal was strong enough to clearly show the "dance" of the arms and legs. The radar could distinguish the human motion even though the person was hidden around the corner.

The Results and the Catch

The Good News:
The experiment proved that using this "Smart Mirror" allows radar to detect human movement around corners with much higher clarity. It turns a weak, confusing whisper into a clear shout. This is huge for security (seeing intruders around corners) and self-driving cars (spotting pedestrians before they step into the road).

The Catch (Limitations):
The "mirror" isn't perfect yet.

Distance: The mirror works best when the person is relatively close. As the person walks further away, the signal gets weaker again.
Field of View: The mirror has a specific angle it can cover. If the person moves too far to the left or right, the mirror can't focus the beam on them anymore.
Narrow Bandwidth: The current test used a specific, narrow slice of radio frequencies. Future versions need to handle a wider range of frequencies to get even better pictures.

The Bottom Line

This paper shows that by adding a programmable "smart mirror" to our radar systems, we can finally see what's happening around corners with high precision. It's a step toward making our cities and cars safer by giving them "super-vision" that doesn't require a direct line of sight.

1. Problem Statement

The paper addresses the challenges of Around-the-Corner (ACR) radar sensing in Non-Line-of-Sight (NLOS) environments, which are critical for security, surveillance, and automotive look-ahead warning systems.

Current Limitations: Traditional ACR sensing relies on multipath signals bouncing off walls to detect targets hidden around corners. However, this approach suffers from:
- Weak Signal Strength: Multipath signals undergo significant attenuation, leading to poor detection metrics.
- Localization Ambiguity: Multiple bounces create multiple possible target locations, violating uniqueness criteria and reducing localization accuracy.
- Frequency Constraints: At higher frequencies (above X-band), signals cannot penetrate walls and must rely on surface scattering, exacerbating signal loss.
The Gap: While Reconfigurable Intelligent Surfaces (RIS) have been theoretically proposed to solve these issues, there is a lack of experimental validation using real-world measurement data, particularly regarding practical constraints like limited bandwidth and beamwidth.

2. Methodology

The authors developed an experimental setup to validate RIS-aided ACR sensing at 5.5 GHz.

Experimental Environment:
- An L-shaped corridor (2.4 m wide, 30 cm brick walls) was used.
- Radar: A monostatic radar system configured using a FieldFox N9926A Vector Network Analyzer (VNA) with two HF907 horn antennas.
  - Frequency: Centered at 5.5 GHz with a 10 kHz bandwidth.
  - Sampling: 370 Hz sampling rate to capture Doppler shifts without aliasing.
- Target: A human walking at an average speed of 1.5 m/s.
- Scenarios: Four trajectories were tested:
  - T1 & T3: Line-of-Sight (LOS) motion (tangential and radial).
  - T2 & T4: NLOS motion (tangential and radial around the corner).
RIS Design and Deployment:
- Hardware: A custom RIS board with dimensions 256 × 160 mm, comprising a 16 × 10 array of unit cells.
- Technology: 1-bit coding metasurface unit cells loaded with PIN diodes, providing a fractional bandwidth of 18.51% centered at 5.45 GHz.
- Configuration: The RIS was placed 1.8 m from the radar at an incidence angle of -60° relative to the negative x-axis. It was tuned to reflect beams toward the NLOS region.
- Beam Steering: The RIS was tested at three different reflection angles: 30°, 45°, and 60°.
Signal Processing:
- Time-domain scattering parameters ( $S_{21}$ ) were collected.
- Short-Time Fourier Transform (STFT) was applied to generate micro-Doppler spectrograms to analyze human motion signatures (torso, arms, legs).

3. Key Contributions

Experimental Validation: Unlike previous theoretical studies, this work provides real-world measurement data demonstrating RIS efficacy in ACR scenarios.
Micro-Doppler Recovery: The study successfully demonstrated that RIS can recover micro-Doppler signatures (specifically the motion of walking humans) in NLOS conditions where traditional multipath sensing fails to produce detectable patterns.
Performance Comparison: The paper quantitatively compares three scenarios:
1. Free Space (LOS).
2. Traditional ACR (Multipath only, no RIS).
3. RIS-Aided ACR.
Practical Insights: It highlights the trade-offs of RIS deployment, specifically the impact of two-way transmission path loss and the limitations imposed by the RIS's field of view (FoV).

4. Results

Baseline (No RIS):
- LOS: Strong returns with clear micro-Doppler signatures.
- NLOS (Multipath only): Signals were extremely weak. Micro-Doppler patterns were indistinguishable, making human movement detection around corners nearly impossible.
RIS-Aided NLOS:
- Detection: The introduction of the RIS significantly enhanced signal strength in NLOS regions. Clear micro-Doppler signatures of walking humans were captured.
- Angle Sensitivity: The RIS was effective at reflection angles of 30°, 45°, and 60°. The strongest returns were observed when the target was closer to the RIS.
- Signal Characteristics: While the signal strength was weaker than in free space (due to two-way path loss through the RIS), it was more robust than the traditional multipath-only NLOS case.
- Artifacts: Some Doppler spreading was observed due to residual multipath reflections from walls, but the primary target signature remained distinct.
Distance Impact: As the target moved further from the RIS (e.g., from 2m to 3m), the reflected power decreased, indicating that the current RIS design has a limited effective field of view for long-range detection.

5. Significance and Future Scope

Significance: This research proves that RIS is a viable, low-cost, and energy-efficient solution for enhancing radar sensing in urban NLOS environments. It moves the concept from theoretical simulation to practical hardware implementation, enabling the detection of human activity around corners with high fidelity.
Limitations: The current system is limited by the narrow bandwidth of the radar and the fixed field of view of the specific RIS prototype.
Future Work: The authors propose:
- Expanding the RIS field of view to improve detection at longer ranges.
- Investigating RIS performance with wideband radar signatures.
- Studying scenarios involving multiple targets simultaneously.

In conclusion, the paper establishes that strategically deploying an RIS can overcome the signal attenuation and ambiguity issues inherent in traditional around-the-corner radar, effectively enabling the detection of human micro-motions in NLOS environments.