Abstract
Reconfigurable Intelligent Surfaces (RIS) are pivotal for the realization of 6G wireless networks, yet their reliability is often compromised by the failure of individual meta-atoms. Traditional diagnostic methods are either hardware-intensive or computationally demanding and often fail when a large number of elements are defective. This paper presents a rapid and robust diagnostic framework using space-time-coding (STC) modulation with orthogonal Hadamard codes. By assigning unique temporal signatures to RIS elements, the system can identify faulty components in a single measurement period. Experimental results demonstrate that this method maintains high accuracy even in high-noise environments and scenarios where over 50% of the surface is non-functional.
Introduction
Reconfigurable Intelligent Surfaces (RIS) act as programmable "smart mirrors" that shape electromagnetic environments to enhance wireless coverage and energy efficiency. These surfaces typically consist of hundreds or thousands of sub-wavelength elements, or meta-atoms, controlled by electronic components like PIN diodes. Given the scale and complexity of these deployments, elements frequently fail due to manufacturing defects, environmental degradation, or soldering issues.
Maintaining the performance of an RIS requires efficient diagnostic tools. However, current approaches face significant hurdles. Hardware-based methods, such as near-field scanning, require expensive specialized equipment and controlled environments like anechoic chambers. Signal-level methods, such as those based on Compressive Sensing (CS), are often computationally expensive and rely on the assumption that faults are "sparse." When an RIS is heavily damaged, these traditional signal-processing methods often break down.
Methodology: Space-Time-Coding and Hadamard Matrices
To overcome these limitations, this research introduces a diagnostic method based on Space-Time-Coding (STC). The core innovation lies in using orthogonal Hadamard codes to "label" the reflection from each RIS element.
The Coding Strategy
Each individual element (or column) of the RIS is assigned a unique row from a Hadamard matrix. These matrices consist of +1 and -1 values, which map directly to the 1-bit phase states (0 and $\pi$) of the RIS elements. As the RIS is illuminated by a monochromatic wave, the elements toggle their states according to their assigned temporal codes.
Decoding and Fault Identification
At the receiver, the reflected signal is processed using a simple correlation with the known Hadamard codes. Because the codes are orthogonal, the power contributed by each element is recovered into independent "code channels."
Fault identification is then reduced to a simple power thresholding problem:
- Healthy Elements: Show high power in their assigned channel because their modulation matches the expected code.
- Faulty Elements: Elements that fail to modulate (remaining in a static state) result in near-zero power in their assigned channel, as their static reflection does not correlate with the time-varying Hadamard signature.
To scale this to larger arrays, the researchers implemented a segmented coding scheme, modulating subsets of the array at a time to reduce required code lengths and mitigate mutual coupling between adjacent elements.
Experimental Validation
The method was tested using both numerical simulations and physical prototypes:
- Prototype 1: A 16×16 column-controlled programmable metasurface operating at 9.8 GHz.
- Prototype 2: An 8×8 column-controlled programmable metasurface operating at 10 GHz.
The researchers evaluated 21 different fault scenarios, including extreme cases where more than half of the elements were intentionally disabled.
Key Insights
Sparsity Independence
Unlike Compressive Sensing, which requires faults to be rare to be detectable, the STC method is just as accurate when 90% of the board is broken as when 1% is. The orthogonal nature of the codes ensures that energy from a static (broken) element does not interfere with the diagnosis of healthy ones.
Robustness to Non-Ideal Hardware
In real-world conditions, RIS elements rarely provide a perfect 180° phase shift. The study demonstrated that even with significant phase deviations (e.g., 160° or 200°) and amplitude imbalances, the power gap between "normal" and "faulty" elements remained between 10dB and 20dB, allowing for unambiguous detection.
Angle Agnosticism
The diagnostic remains effective across a wide range of incident and receiving angles (up to ±60°). This is a critical feature for field testing, where precise alignment between the transmitter, RIS, and receiver is often difficult to achieve.
Results and Performance
Simulation data showed that the diagnostic error rate (ER) drops below $10^{-4}$ once the Signal-to-Noise Ratio (SNR) exceeds 15dB. In experimental trials, the method successfully identified faults with high contrast, requiring only a single data collection period and minimal signal processing power. This represents a significant efficiency gain over iterative algorithms.
Limitations and Future Work
While the method is highly effective, the authors noted a power imbalance (up to 5dB) between central and edge columns in larger prototypes, likely due to the incident wave not being a perfect plane wave. Future refinements may involve power-leveling normalization for very large surfaces. Additionally, while the current study focused on 1-bit RIS, the authors suggest that Hadamard codes could be adapted for multi-bit configurations in future research.
Conclusion
This space-time-coding diagnostic approach provides a fast, computationally light, and highly robust solution for RIS maintenance. By moving away from complex iterative reconstructions and toward simple orthogonal correlation, the method enables reliable "single-shot" diagnostics that are practical for real-world 6G deployments.
Summary updated