An Experimental Validation of Reconfigurable Intelligent Surfaces Achieving Pulse Width-Modulated Singular Reflection Angles Without External Power Sources

TL;DR

The paper tackles the challenge of RIS complexity and synchronization by introducing a passive, waveform-selective metasurface that autonomously toggles between two singular reflection angles as the incident pulse width changes. It adopts a diode-bridge, capacitor-based nonlinear circuit integrated into subwavelength unit cells to create SP and CW–dependent reflection states, assembling them into a 10-unit-cell supercell to generate a controlled phase gradient. Through both simplified impedance-sheet simulations and full EM modeling, plus a 3 GHz experimental prototype, the authors demonstrate specular reflection at $\theta_{r1}$ for SP and an anomalous reflection at $\theta_{r2}$ for CW, with a $2\pi$ phase gradient under CW. When combined with BPSK, the RIS yields distinct signal strengths at two receivers and BER improvements of over $7$ dB in favorable configurations, indicating practical opportunities for pulse-width–dependent RIS in next-generation wireless networks while avoiding external power or line-based control. Overall, the work suggests a path to simpler, self-adaptive wireless environments using passive waveform-sensitive metasurfaces that can support dual-beam operation without centralized control.

Abstract

In this study, we introduce a design concept that leverages pulse width variation to enable a reconfigurable intelligent surface (RIS) and to autonomously switch reflection properties between two angles without any active control system. Our RIS alters its beam pattern from a singular specular reflection to another unique singular anomalous reflection when the incoming waveform changes from a short pulse to a continuous wave, even at the same frequency. Unlike conventional RISs, our passive control mechanism eliminates the requirements of active components and precise symbol-level synchronization with the transmitting antennas, reducing the system complexity level while offering dynamic material adaptability. We numerically show that the proposed RIS design is capable of varying the received magnitude of an incident wave by a factor of ten, which is also experimentally validated for the first time. Employing binary phase-shift keying (BPSK) modulation, we further report that the communication characteristics can be varied by 7 dB or more, which indicates that the proposed design is not limited to a single frequency component as long as the bandwidth of the given signal is covered by that of the RIS design. These results may present new opportunities for exploring and deploying pulse width-dependent RISs in practical scenarios involving next-generation communication systems.

Figures (12)

• Figure 1: A visualization of the proposed RIS and its ability to manipulate the beamforming process based on the pulse widths of the incoming wireless signals without an external power source (e.g., a direct-current (DC) supply) as a passive solution to outdoor communications. The incident angle $\theta_{in}$ and reflected angles $\theta_{r1}$ and $\theta_{r2}$ can be customized by the RIS design.
• Figure 2: The periodic unit cell configuration. (a) A unit cell configuration consisting of metallic patches (gray) on top of a grounded dielectric substrate (yellow). (b) A circuit configuration with a connection between the metallic patches. The detailed geometrical and circuit parameters are shown in Tables \ref{['tab:1']} to \ref{['tab:3']}.
• Figure 3: The reflecting profiles of the periodic unit cell shown in Fig. \ref{['fig:2']}. (a) The reflection magnitudes and (b) phases for the 50-ns SPs and CWs.
• Figure 4: The reflection profiles produced for the unit cells at 3 GHz with various $L_{add}$ values (see Table \ref{['tab:4']} for the specific inductance values).
• Figure 5: (a). The simplified RIS model based on homogeneous impedance sheets representing the impedances of UC$\#$1 to UC$\#$10. The geometric parameters are given in Table \ref{['tab:5']}. The equivalent surface impedances, represented by $Z_1$ to $Z_{10}$, are shown in Table \ref{['tab:6']}. (b) The simulated reflection profiles at a normal incidence.