Array-Fed RIS: Validation of Friis-Based Modeling Using Full-Wave Simulations

TL;DR

This work addresses validating a Friis-based propagation model for an array-fed RIS architecture (AMAF-RIS) by comparing a simplified transmission matrix $T_{Friis}$ to a full-wave CST solution $T_{full}$. The authors design a 150 GHz patch-based RIS/AMAF system (16×16 RIS, 2×2 AMAF) and extract $T_{full}$ from S-parameters to assess the accuracy of the Friis model for system design. Key findings show the Friis model yields a larger top singular value (e.g., $T_{Friis}$ vs $T_{full}$) and a substantially larger RIS taper (29.5 dB vs 18.1 dB) with similar beam directions, while full-wave results are more sensitive to beam pointing errors; rate-CDFs are qualitatively aligned. The study concludes that Friis-based models are suitable for initial design, whereas full-wave simulations are essential for final performance evaluation, and it points to future work on stacked AMAF-RIS configurations, dual polarization, and hardware validation.

Abstract

Space-fed large antenna arrays offer superior efficiency, simplicity, and reductions in size, weight, power, and cost (SWaP-C) compared to constrained-feed systems. Historically, horn antennas have been used for space feeding, but they suffer from limitations such as bulky designs, low aperture efficiency ($\approx 50\%$), and restricted degrees of freedom at the continuous aperture. In contrast, planar patch arrays achieve significantly higher aperture efficiency ($>90\%$) due to their more uniform aperture distribution, reduced weight, and increased degrees of freedom from the discretized aperture. Building on these advantages, we proposed an array-fed Reflective Intelligent Surface (RIS) system, where an active multi-antenna feeder (AMAF) optimizes power transfer by aligning with the principal eigenmode of the AMAF-RIS propagation matrix $\mathbf{T}$. While our previous studies relied on the Friis transmission formula for system modeling, we now validate this approach through full-wave simulations in CST Microwave Studio. By comparing the Friis-based matrix, $\mathbf{T}_{\rm Friis}$, with the full-wave solution, $\mathbf{T}_{\rm full. wave}$, we validate the relevance of the Friis-based modeling for top-level system design. Our findings confirm the feasibility of the proposed AMAF-RIS architecture for next-generation communication systems.

Figures (8)

• Figure 1: Patch Element connected to a discrete port highlighted in red. The position of the discrete port is shifted $\unit[80]{\mu m}$ from the patch center to ensure a base impedance close to $\unit[50]{\Omega}$.
• Figure 2: Reflection at individual patch elements (4 AMAF elements).
• Figure 3: The 3D AMAF-RIS model in CST Microwave Studio®. The AMAF with $N_a=4$ patches is visible from the back and the RIS with $N_p = 256$ patches is visible from the front.
• Figure 4: Mutual coupling between AMAF elements.
• Figure 5: RIS PEM magnitude taper (normalized) as obtained from full-wave simulations.