Source-linked AI summary
Design and Evaluation of Reconfigurable Intelligent Surfaces in Real-World Environment
Georgios C. Trichopoulos, Panagiotis Theofanopoulos, Bharath Kashyap, Aditya Shekhawat, Anuj Modi, Tawfik Osman, Sanjay Kumar, Anand Sengar, Arkajyoti Chang, Ahmed Alkhateeb
TL;DR
Prior RIS evidence was concentrated in simulations or constrained prototypes, leaving real-world performance insufficiently evaluated. This paper develops and tests a low-power RIS in realistic outdoor communication settings, reporting SNR gains and coverage improvement, while noting that the prototype is not optimized.
Problem
Prior RIS performance studies were mainly limited to simulation data, motivating real-world prototype-based evaluation.
Method
The paper develops a proof-of-concept RIS and evaluates its beamforming and coverage gains in realistic outdoor settings.
Results
The developed RIS provides around 20 dB SNR gain with directional antennas and achieves 8 dB SNR gain with an omni-directional antenna.
Takeaways & Limitations
RISs can extend wireless coverage in strongly occluded scenarios, including when the surface is in the far field of both base station and user.
Takeaways & Limitations
The current prototype is non-optimized.
Abstract
from arXiv · showhide
Reconfigurable intelligent surfaces (RISs) have promising coverage and data rate gains for wireless communication systems in 5G and beyond. Prior work has mainly focused on analyzing the performance of these surfaces using computer simulations or lab-level prototypes. To draw accurate insights about the actual performance of these systems, this paper develops an RIS proof-of-concept prototype and extensively evaluates its potential gains in the field and under realistic wireless communication settings. In particular, a 160-element reconfigurable surface, operating at a 5.8GHz band, is first designed, fabricated, and accurately measured in the anechoic chamber. This surface is then integrated into a wireless communication system and the beamforming gains, path-loss, and coverage improvements are evaluated in realistic outdoor communication scenarios. When both the transmitter and receiver employ directional antennas and with 5m and 10m distances between the transmitter-RIS and RIS-receiver, the developed RIS achieves $15$-$20$dB gain in the signal-to-noise ratio (SNR) in a range of $\pm60^\circ$ beamforming angles. In terms of coverage, and considering a far-field experiment with a blockage between a base station and a grid of mobile users and with an average distance of $35m$ between base station (BS) and the user (through the RIS), the RIS provides an average SNR improvement of $6$dB (max $8$dB) within an area $> 75$m$^2$. Thanks to the scalable RIS design, these SNR gains can be directly increased with larger RIS areas. For example, a 1,600-element RIS with the same design is expected to provide around $26$dB SNR gain for a similar deployment. These results, among others, draw useful insights into the design and performance of RIS systems and provide an important proof for their potential gains in real-world far-field wireless communication environments.
I. INTRODUCTION
RISs use tunable reflective structures to redirect or focus wireless signals, addressing propagation and coverage challenges. This work motivates real-world evaluation through a low-power, portable RIS prototype.
- RISs offer potential coverage improvements by intelligently reflecting signals toward receivers, particularly for non-line-of-sight links.Anomalous reflection can create viable propagation paths when users are outside the specular direction.
- The paper develops a low-power, portable proof-of-concept prototype to validate RIS gains in realistic communication environments.The motivation is to evaluate RIS performance in reality rather than rely only on analytical or simulated expectations.
- RIS unit cells modulate re-radiated amplitude and phase through tunable electromagnetic properties, such as switched patch-antenna terminations.Additional switches can enable multi-bit wavefront modulation for improved beam control and efficiency.
- RISs can redirect far-field signals or focus near-field energy by changing the amplitude and phase of reflected waves.Their subwavelength unit cells receive incident signals and re-radiate them with configurable properties.
- A bistatic radar-equation model estimates received power when an RIS is placed between a base station and user equipment.The model incorporates transmit power, antenna gains, propagation distances, incidence and departure angles, RIS area, efficiency, and wavelength.
- 20 dB stronger signal results from a ten-fold increase in RIS area under the same propagation scenario.Received power increases quadratically with RIS size and inversely with the square of wavelength.
B. Prior Work on Reconfigurable Intelligent Surfaces
Prior RIS research spans hardware, beamforming, channel acquisition, and coverage analysis, but much of the evidence remains simulation-based or constrained to near-field settings. The paper targets scalable, realistic far-field evaluation.
- Channel acquisition creates high training overhead, especially for nearly passive RISs.Prior work proposed sparse active elements and learning-based approaches to help estimate RIS channels.
- Prior RIS studies on design, beamforming, and coverage were limited to simulation data.This motivates proof-of-concept prototypes and real-world assessments of coverage and data-rate gains.
- Existing experiments included near-field indoor coverage and configurations requiring directive antennas for both transmitter and receiver.These settings do not establish performance when both communication terminals are in the RIS far field.
- A prior RIS prototype used varactor switches, sub-λ/2 unit cells, vias, and a multilayer PCB structure.The cited work describes this as viable at microwave frequencies but non-scalable for mmWave and THz frequencies.
- Higher-frequency RIS implementations require new switching materials and simplified unit-cell layouts for practical manufacturability.Examples include VO2, graphene, and liquid crystal approaches intended to overcome PIN- and varactor-diode limitations.
- RIS systems face engineering challenges involving switching performance and biasing-circuit manufacturability.Practical surfaces are also expected to scale to large areas, use low power, and fit varied surfaces.
C. Contribution
The paper contributes a scalable 5.8 GHz RIS design and evaluates its beamforming and coverage performance in realistic outdoor environments, including obstructed links and far-field settings.
- The study designs and fabricates a low-power RIS, then evaluates beamforming gains in realistic near-field and far-field outdoor settings.Measurements account for scattering from the ground and surrounding environment.
- The single-layer, single-switch-per-cell design operates at 5.8 GHz and is compatible with mmWave and THz fabrication technologies.It requires no vias or same-plane integrated biasing network and supports electronic beam scanning in azimuth and elevation.
- The work characterizes RIS beamforming through anechoic-chamber radiation measurements and outdoor testing.The characterization includes unit-cell, switch, beamforming, and control-circuit integration aspects.
- The coverage experiments use mobile UEs with omni-directional antennas in outdoor areas where a building blocks the BS-to-UE line of sight.The RIS provides an alternative signal path beyond direct LoS coverage.
- The wireless testbed evaluates SNR improvement for mobile users and coverage extension beyond line-of-sight areas.It includes a base station, user equipment, and the RIS in realistic field tests.
II. RIS-BASED WIRELESS COMMUNICATION SYSTEM
The section formulates RIS-assisted communication and beamforming, representing each RIS element by a controllable phase and designing configurations for desired incident and reflection directions. A 1-bit codebook discretizes these phases into two switch states for practical operation.
- System model: The system models a single-antenna transmitter and receiver communicating through a reconfigurable intelligent surface.
- RIS representation: The RIS reflection vector contains one phase-control coefficient per element, with element mn represented as e^jϕ_mn.
- Beamforming objective: The beamforming vector is designed to maximize achievable rate, neglecting the direct line-of-sight link in blocked scenarios.
- 1-bit beamforming codebook: A single-bit quantization scheme rounds RIS phases to 0° or 180°, implemented as OFF and ON PIN-diode states.The scheme is adopted for simplicity and lower cost than higher-bit quantization.
- Beamforming codebook: The RIS codebook contains predefined phase configurations that redirect waves from selected incident directions to selected reflection directions.For incident-direction set I and reflection-direction set D, the codebook has N_iN_d codewords.
III. RECONFIGURABLE INTELLIGENT SURFACE DESIGN
The RIS is designed as a 5.8 GHz, single-layer structure with integrated PIN-diode switching and control circuitry. Its topology avoids vertical connections while enabling phase modulation through the unit-cell design.
- Prototype design: The prototype design includes the unit cell, fixed- and multi-beam RIS measurements, and integrated control circuitry.
- Operating frequency: The RIS operates at 5.8 GHz in the unlicensed spectrum used for wireless local area networks.
- Topology: The proposed topology uses a single layer besides the ground plane and requires no vertical components such as vias.This structure is intended to reduce losses and fabrication complexity relevant to future mmWave and THz RISs.
- Unit-cell design: Each unit cell uses a main resonant patch connected to a parasitic patch through an RF PIN diode and biasing lines.
- Switching mechanism: Switch activation changes the reflected signal phase by 180° without significant magnitude modulation within the operating bandwidth.
KEY DIMENSIONS OF THE RIS UNIT CELL
The RIS uses a 160-element planar array with a single-switch, no-via unit-cell topology. Electromagnetic analysis predicts 180° switching with limited magnitude variation, while quantization sidelobes constrain scanning in this implementation.
- Array dimensions: The RIS comprises 160 unit cells containing passive antennas, RF PIN switches, and biasing lines.
- Topology: The proposed topology uses one tuning switch per unit cell and a single layer without vertical ground connections.
- Sidelobe limitation: Phase quantization produces a second grating lobe near the opposite angle around the specular direction.Random phase delays are identified as a possible mitigation, but sidelobe elimination is left for future prototype implementations.
- Electromagnetic analysis: The unit-cell response is analyzed with periodic-boundary simulations that omit radiation effects at the aperture edge.
- Unit-cell response: At 5.8 GHz, the calculated two-state response has less than 1 dB magnitude modulation and a 180° phase difference.
- Bandwidth: The expected bandwidth for a 180° ± 20° phase-modulation range is approximately 150 MHz.
B. Array Design
The array is evaluated analytically, through full-wave simulations, and in an anechoic chamber using near-field and plane-wave illumination. Simulated and measured patterns agree for the fixed-beam 5.8 GHz prototype, while phase quantization introduces additional lobes.
- B. Array Design: Analytical and full-wave models evaluate RIS radiation under feed-horn illumination at an incidence angle of −27.5° and reflection angles of 0°, 17°, and 60°.
- B. Array Design: Full-wave and analytical results produce distinct main lobes at the desired reflection directions.
- B. Array Design: Differences in sidelobe levels are attributed to diffraction and surface-wave phenomena omitted from the array-factor analysis.
- B. Array Design: Plane-wave simulations evaluate normalized radar cross-section patterns for reflection angles of 22.5°, 40°, and 60°.
- B. Array Design: Phase quantization produces a second grating lobe near the opposite angle around the specular direction.
- C. Characterization of a fixed beam RIS: The fixed-beam prototype contains 160 unit cells and is characterized at 5.8 GHz in an anechoic chamber.
- C. Characterization of a fixed beam RIS: Measured normalized gain over the −90° to +90° E-plane agrees well with full-wave numerical analysis and the designed beam.
D. Implementation of the Reconfigurable Intelligence Surface
The paper implements a 160-element, 5.8 GHz RIS using PIN-diode unit cells, 1-bit phase control, and integrated circuitry. Anechoic-chamber measurements confirm beamforming agreement with simulations and scanning across ±60°.
- Fabrication and control: 188° phase difference is measured between diode biasing states across a wide frequency range around 5.8 GHz.The measured phase difference agrees with circuit simulations and approximates the intended 180° shift.
- Fabrication and control: 160 PIN diodes are assembled into the RIS array and integrated with control circuitry for beam steering.The array uses a 160-bit codeword, with each unit cell producing a 0° or 180° phase delay.
- Fabrication and control: Less than 0.4 Watts of DC power is consumed by the RIS, including its biasing circuitry.The paper notes that transistor-based switches could reduce consumption to microwatt levels in larger RISs.
- Anechoic-chamber characterization: Measured radiation patterns for five reflected directions show very good agreement with full-wave simulations.The tested directions are 0°, 15°, 30°, 45°, and 60°; quantized phase distributions are also reported.
- Anechoic-chamber characterization: The RIS scans across ±60° while maintaining a single main lobe.The side-lobe level remains below -7 dB over 5.75–5.85 GHz and below -5 dB over more than 400 MHz at θd = 30°.
IV. INTEGRATING THE RIS INTO A WIRELESS COMMUNICATION TESTBED
The fabricated RIS is integrated into a 5.8 GHz OFDM wireless testbed with single-antenna transmitter and receiver hardware. Beam training switches through RIS codewords and selects the configuration producing the strongest received signal.
- Beam selection: The RIS codebook configures reflection phases to redirect the transmitted signal toward the receiver.The adopted codebook uses 1-bit phase control, and the controller manages RIS configuration switching.
- Testbed architecture: The communication system uses OFDM with 64 subcarriers, a 5.8 GHz center frequency, and 20 MHz bandwidth.
- Beam selection: A beam-training controller cycles through codebook beams while the transmitter sends OFDM pilot sequences.The receiver measures received power on each subcarrier and identifies the best RIS configuration.
V. FIELD TESTS AND RESULTS
Outdoor field campaigns evaluate RIS beamforming and coverage under realistic propagation conditions, including scattering, edge diffraction, and line-of-sight blockage. Two setups measure directional steering and blocked-user coverage.
- Field-test design: All measurement campaigns are conducted outdoors to capture real-world propagation phenomena such as terrain scattering and edge diffraction.The authors frame these measurements as providing insights into realistic wireless deployments.
- Measurement Setup 1: The parking-lot receiver is rotated from 0° to 60° at 10 m, with additional path-loss measurements at 20 m and 40 m.
- Measurement Setup 2: The auditorium setup places a concrete wall between the base station and receiver to emulate severe signal blockage.The receiver uses an omnidirectional antenna and moves across the area near the venue entrance.
- Measurement Setup 2: The blocked-user experiment compares received signal levels with and without RIS beam scanning across a measurement grid.The RIS is positioned in line of sight to both the base station and the occluded user region.
B. Characterization of the Beamforming Codebook
Field measurements confirm that the RIS steers beams toward the receiver and improves received SNR in directional and blocked-link scenarios. Path loss increases with distance and with steering away from broadside.
- Beamforming performance: The measured peak power occurs at the receiver location corresponding to each scanned RIS beam direction.This confirms agreement between codebook beam directions and the receiver position along the circular path.
- Beamforming performance: Coupling appears for receiver angles below 20° because of transmitter backlobes and close transmitter–receiver proximity.The paper identifies this as a measurement-setup artifact that would not occur when the transmitter is farther away or blocked.
- Beamforming performance: 18–20 dB SNR gain is achieved in the directional field deployment.The gain is observed while scanning RIS beams over the considered receiver directions.
- Path loss: Path loss increases as receiver distance increases and as the beam deviates from broadside.Measurements compare nominal steering angles of 10°, 20°, and 30° at distances up to 40 m.
- Coverage improvement: 6 dB average SNR improvement and up to 8 dB maximum improvement are measured in the blocked region.The coverage experiment uses a 28-point grid, with the RIS redirecting the base-station signal around the concrete-wall occlusion.
VI. CONCLUSION
The paper demonstrates RIS beamforming and coverage gains in realistic wireless environments using a scalable, low-power 160-element prototype. It identifies practical design strengths, energy benefits, and a remaining beamforming limitation for future refinement.
- Evaluation: The prototype was integrated into a wireless communication system to evaluate beamforming and coverage gains in realistic communication scenarios.Beamforming was characterized for both passive and active implementations.
- Prototype design: A 160-element, sub-6GHz RIS prototype was designed and fabricated as a scalable single-layer reflectarray supporting azimuth and elevation scanning.Its simple geometry and single-bit switches support scaling toward mmWave and THz implementations.
- Measured gains: Around 20dB SNR gain was achieved with directional transmitter and receiver antennas at 5m and 10m RIS distances.The reported configuration demonstrates substantial beamforming gain under directional-antenna conditions.
- Measured gains: 8 dB SNR gain was achievable over a 35m average BS-RIS-UE path with a blocked LoS link and one directional and one omni-directional antenna.The gain is reported for a far-field coverage scenario involving asymmetric antenna directivity.
- Scalability: A ten-fold RIS-area increase to a 1,600-element array was reported to further increase SNR by 20.The result supports direct scalability of the surface area, although the passage does not specify the unit for the final increase.
- Power and future design: The prototype consumes less than 0.4 Watt including biasing circuitry, while future designs could reduce power further using transistor-based switches.A cited alternative reports less than 4 uWatt of DC power for an RIS with 10,000 switches.
- Power and future design: Future beamforming designs should eliminate grating lobes in the opposite direction of the main beam and mitigate low-bit quantization errors with fixed random phase delays.These changes target beamforming artifacts associated with the current design.