Wideband Reconfigurable Reflective Metasurface with 1-Bit Phase Control Based on Polarization Rotation

Abstract

The rapid expansion of broadband wireless communication systems, including 5G, satellite networks, and next-generation IoT platforms, has created a strong demand for antenna architectures capable of real-time beam control, compact integration, and broad frequency coverage. Traditional reflectarrays, while effective for narrowband applications, often face inherent limitations such as fixed beam direction, high insertion loss, and complex phase-shifting networks, making them less viable for modern adaptive and reconfigurable systems. Addressing these challenges, this work presents a novel wideband planar metasurface that operates as a polarization rotation reflective metasurface (PRRM), combining 90° polarization conversion with 1-bit reconfigurable phase modulation. The metasurface employs a mirror-symmetric unit cell structure, incorporating a cross-shaped patch with fan-shaped stub loading and integrated PIN diodes, connected through vertical interconnect accesses (VIAs). This design enables stable binary phase control with minimal loss across a significantly wide frequency range. Full-wave electromagnetic simulations confirm that the proposed unit cell maintains consistent cross-polarized reflection performance and phase switching from 3.83 GHz to 15.06 GHz, achieving a remarkable fractional bandwidth of 118.89%. To verify its applicability, the full-wave simulation analysis of a 16 × 16 array was conducted, demonstrating dynamic two-dimensional beam steering up to ±60° and maintaining a 3 dB gain bandwidth of 55.3%. These results establish the metasurface’s suitability for advanced beamforming, making it a strong candidate for compact, electronically reconfigurable antennas in high-speed wireless communication, radar imaging, and sensing systems.

1. Introduction

Reconfigurable reflective metasurfaces (RRMs) have emerged as a transformative technology for electromagnetic wave manipulation, offering lightweight, low-cost, and energy-efficient alternatives to traditional phased arrays [1,2,3]. By utilizing spatial feeding architectures, RRMs eliminate the reliance on bulky transmit/receive (T/R) modules typically used in conventional phased arrays. This architectural simplification results in a substantial reduction in insertion loss compared to traditional feed network configurations [4,5]. Despite these advantages, achieving wideband operation remains a critical challenge. Most RRMs are limited to narrowband or single-frequency performance due to the inherent bandwidth constraints of their unit cells [6], restricting their applicability in next-generation wireless communications, radar systems, and other broadband applications. Recent research has focused on overcoming this limitation through simplified design approaches, such as 1-bit phase quantization, to balance performance and practicality [7,8,9,10].

A key advancement in RRM design is the use of polarization rotation techniques, which exploit cross-polarization conversion to achieve broadband phase control [11]. Since RRAs depend on their constituent elements, the reflectarray’s capabilities are ultimately constrained by element performance. The required phase progression is conventionally implemented using resonant unit cell architectures [12], where traditional resonant-based unit cells achieve phase control through active components, such as PIN diodes. However, their phase responses are typically frequency-dependent, resulting in narrow bandwidths [13], and deviations from the resonant frequency often lead to significant phase errors and amplitude degradation, further restricting operational bandwidth [14]. While multilayer substrates have been employed to enhance bandwidth [15], the strong coupling between phase and amplitude in resonant designs poses a persistent challenge, making it difficult to achieve both stable phase control and high reflection efficiency over a broad frequency range. Consequently, wideband RRM optimization hinges on two objectives: (i) phase response stability across frequency, and (ii) reflection efficiency preservation throughout the operational band. Polarization state control has emerged as an essential technique, permitting metasurfaces to realize stable ${180}^{\circ }$ phase differences (for 1-bit operation) across the target frequency band. Among available strategies, polarization state control has gained attention as an effective method to ensure phase stability and broaden bandwidth performance. Polarization rotation elements are categorized as either transmissive [16] or reflective [17] configurations. These typically utilize split-ring resonator variants employing standard, double [18,19], triangular [20], or L-shaped [21] or substrate-integrated waveguide (SIW) cavities [22]. Recent research has focused on active RRMs for 1-bit operation, where unit cells integrate radiating patches, PIN diodes, and VIAs to enable dynamic beamforming [23,24,25,26]. All designs share a fundamental symmetry characteristic, typically exhibiting 45° rotational symmetry to enable interaction with both vertically and horizontally polarized linear waves.

Significant efforts have been made to overcome bandwidth limitations in reconfigurable metasurfaces, yet fundamental challenges persist. Ref. [9] demonstrated a 1-bit circularly polarized (CP) magneto-electric dipole with 44% fractional bandwidth (8.3–13 GHz) using balanced dipole responses, enabling beam scanning up to $\pm {45}^{\circ }$ via PIN diodes. However, this design suffers from inherent 1-bit quantization losses, which degrade beamforming efficiency. The study in [10] presents a design that incorporates a PIN diode within a cross bow-tie patch, augmented with gradient-shaped patches and a parasitic shunting structure that effectively suppresses diode loss by limiting current flow through the switching device. Through parametric optimization, the authors achieved a maximum fractional bandwidth of 50% (6.75–11.25 GHz). While this work demonstrates a significant improvement in bandwidth, the element configuration remains relatively complex due to the multilayered parasitic design, which may introduce fabrication challenges and increase sensitivity to assembly tolerances. Further advancements include [11], which reported a wideband, beam-steerable reflectarray antenna with 33% fractional bandwidth (8.3–11.5 GHz) using polarization-rotating phase shifters. Similarly, the work in [24] demonstrated a polarization rotation-based element achieving a 17% bandwidth (11.8–14 GHz) by introducing a bow-tie slot in the ground plane, effectively preserving performance despite diode-induced symmetry breaking. Although this approach is innovative, it highlights a typical trade-off in such designs—enhancing reconfigurability often results in constrained operational bandwidth. Ref. [27] pushed the bandwidth further to 41.9% (6.82–9.96 GHz) using a multi-resonant unit cell with varactor diodes for dynamic phase control. This design incorporates anisotropic meta-atoms to enable polarization rotation and beam scanning but requires complex biasing networks. Ref. [28] achieved an ultra-broadband 45.1% fractional bandwidth (6.7–10.6 GHz) by combining mirror-symmetry-enabled 1-bit phase control with sub-wavelength engineering, supporting multipolarized operation and wide-angle beam steering. Ref. [29] demonstrated a reconfigurable unit cell with 40% bandwidth (10.4–15.7 GHz) across X/Ku bands using a four-patch structure and optimized PIN diode placement, though it struggles with polarization isolation at band edges. These prior approaches face inherent constraints that limit their practical implementation. Conventional designs typically trade bandwidth for reconfigurability, with most demonstrations constrained to less than 50% fractional bandwidth due to their reliance on resonant structures. Fabrication challenges arise in VIA-dependent and multi-resonant designs, where sub-wavelength precision requirements increase manufacturing complexity. Furthermore, existing polarization-rotating metasurfaces often exhibit performance degradation at band edges due to dispersion mismatches between orthogonal polarization states.

In this work, a novel reconfigurable reflective metasurface element is proposed that combines 1-bit phase modulation and broadband polarization rotation within a compact structure. The key innovations of the design include: (i) 1-bit binary phase switching based on polarization rotation rather than conventional resonance phase shifting; (ii) ultra-broadband operation from 3.83 to 15.06 GHz, covering S, C, X, and part of the Ku-band with a fractional bandwidth of 118.89%; (iii) maintenance of high polarization conversion efficiency, with cross-polarized reflection levels exceeding –1 dB across the band; and (iv) realization of two-dimensional (2D) beam steering up to ±60° with a stable 3 dB gain bandwidth of 55.3%. In contrast to previously reported designs, such as [9] with 44% fractional bandwidth, [10] with 50%, [11] with 33%, and [28] with 45.1%, the proposed structure demonstrates a notable improvement in frequency coverage and reconfigurability. Furthermore, the metasurface adopts a simpler biasing and structural configuration while preserving consistent performance across the entire spectrum. These attributes position the design as a promising solution for future wideband, polarization-tunable, and electronically reconfigurable antenna systems in high-frequency communication platforms.

2. Design, Analysis and Optimization of a 1-Bit RRM Element

2.1. Theory and Principle

The behavior of the proposed reconfigurable reflective metasurface can be effectively described using the polarization reflection matrix formalism, which characterizes how the incident electric field transforms upon reflection from the metasurface surface. For a normally incident, linearly polarized wave propagating along the negative z-direction and impinging on a metasurface situated in the xy-plane, it is governed by the polarization reflection matrix. The relationship between the reflected and incident electric field components can be expressed using a polarization reflection matrix formalism:

$$\left[\begin{array}{c}{E}_x^r\\ E_y^r\end{array}\right] = \left[\begin{array}{cc}{r}_{xx}& r_{xy}\\ r_{yx}& r_{yy}\end{array}\right]\left[\begin{array}{c}{E}_x^i\\ E_y^i\end{array}\right]$$

(1)

In this matrix equation, $E_x^i$ and $E_y^i$ are the x and y components of the incident electric field, respectively, while $E_x^r$ and $E_y^r$ represent the corresponding reflected field components. The matrix

$$R_{LP} = \left[\begin{array}{cc}{r}_{xx}& r_{xy}\\ r_{yx}& r_{yy}\end{array}\right]$$

(2)

encapsulates the metasurface’s linear polarization reflection characteristics. The diagonal elements $r_{xx}$ and $r_{yy}$ quantify the co-polarized reflections, preserving the original polarization, whereas the off-diagonal terms $r_{xy}$ and $r_{yx}$ describe cross-polarized components that indicate polarization rotation or conversion. The metasurface’s reconfigurability is governed by its symmetry properties. Specifically, reflection symmetry about the $yz$-plane introduces a transformation to the polarization response described by a similarity operation on the reflection matrix:

$$R_{LP}^{\prime } = M^{-1}{R}_{LP}M = \left[\begin{array}{cc}{r}_{xx}& -r_{xy}\\ -r_{yx}& r_{yy}\end{array}\right]$$

(3)

This result indicates that the cross-polarized reflection terms $r_{xy}$ and $r_{yx}$ undergo a sign inversion, corresponding to a phase shift of $\pi$, while the co-polarized terms remain unchanged. Such symmetry-induced phase behavior plays a vital role in dynamic polarization control, allowing the metasurface to switch between polarization states by electrically tuning individual RRM elements. By incorporating active components into the reflectarray antenna design, it becomes possible to realize active polarization converters that operate over a broad bandwidth with high polarization conversion rates (PCR). Consequently, the elements can be designed to maintain consistent reflection properties and high polarization conversion rates over an extensive frequency band, generally exhibiting values such that $r_{xx}$ and $r_{yy}$ are close to zero, while $r_{xy}$ and $r_{yx}$ approach unity.

2.2. RRM Element Configuration

The unit cell geometry is based on a central cross-arm configuration augmented with a fan-shaped stub on each arm, designed to support symmetric current distribution and stable reflection behavior. The fan-shaped stub acts as a passive scatterer to enhance bandwidth, as illustrated in Figure 1. The proposed unit cell adopts a cross-shaped patch with fan-shaped stubs based on its inherent ability to support symmetric current distributions and efficient polarization rotation. The mirror-symmetric configuration ensures consistent phase states across the broadband spectrum, while also simplifying biasing through symmetrical diode placement. The fan-shaped extensions introduce additional degrees of freedom to engineer multiple resonances, contributing to the ultra-broad bandwidth. This combination was selected to balance compactness, reconfigurability, and wideband polarization conversion in a single-layer metasurface topology.

The proposed element has physical dimensions of approximately ${\lambda }_0/2$, where ${\lambda }_0$ denotes the free-space wavelength at 10 GHz. The structure consists of three metallic layers separated by two dielectric substrates, both fabricated using F4B material with a relative permittivity of 2.2 and a loss tangent of 0.0022. Substrate 1, located above the ground plane, forms the scattering layer, while Substrate 2, placed below the ground plane, functions as the layer for the direct current (DC) bias network. The two substrates are bonded using a Rogers RO4450F film, which has a dielectric constant of 3.52 and a loss tangent of 0.004, ensuring structural integrity and minimal signal loss as depicted in Figure 1b. A cross-shaped metallic arm integrated with a fan-shaped stub is situated on top of Substrate 1. This configuration includes a central VIA for DC bias connection and four shorting VIAs to the ground plane, enabling it to function as a reconfigurable scatterer under incident electromagnetic waves.

Four PIN diodes are embedded in the gaps between the cross arm and the fan-shaped stub to achieve switching functionality. Two operating states are defined: in the “0” state, diodes D1 and D3 are ON, while D2 and D4 are OFF; in the “1” state, the conditions are reversed. This anti-symmetric arrangement establishes a mirror-symmetry relationship between the two states, which is crucial for dynamic polarization manipulation. The DC bias lines are implemented on the bottom surface of Substrate 2, beneath the ground plane, to avoid interference with the electromagnetic response of the element. All metallic VIAs are designed as through-holes to simplify fabrication. In the simulations, the PIN diodes MADP-000907-14020 are modeled using lumped element representations. For the forward-biased state, a series combination of a 7.8 $\mathrm{\Omega }$ resistance and a 30 pH inductance is used. For the reverse-biased state, a 30 pH inductance in series with a 0.025 pF capacitance is applied. Full-wave electromagnetic analysis is carried out using Ansys HFSS, employing master–slave periodic boundary conditions and Floquet port excitations to evaluate the performance of the designed element.

2.3. Performance Analysis and Theoretical Validation

To validate the effectiveness of the proposed design mechanism, the reflection coefficients $r_{xx}$ and $r_{yy}$ along with their corresponding phase responses were investigated for various cross-arm lengths ($cl$) and cross-arm widths ($cw$), as illustrated in Figure 2. The simulated results shown in Figure 2a present the reflection magnitude characteristics, where it is evident that modifying the cross-arm dimensions significantly affects the fractional bandwidth. Specifically, increasing the cross-arm length ($cl$) from 16 mm to 19 mm progressively shifts the resonance dips to lower frequencies and broadens the bandwidth. Similarly, adjusting the cross-arm width ($cw$) from 1 mm to 2 mm modifies the reflection behavior at higher frequencies, particularly above 10 GHz. Notably, a wider cross-arm $cw = 2 \mathrm{mm}$ results in better impedance matching across a broader frequency range, as indicated by the lower reflection magnitudes below –10 dB between 3.83 and 15.06 GHz. This indicates enhanced absorption or transmission control, enabling improved performance for wideband polarization conversion. Figure 2b shows the reflection phase characteristics corresponding to the same geometrical variations. The phase difference between the orthogonal components remains close to ${180}^{\circ }$ over a wide frequency band when using a longer cross-arm $cl = 19 \mathrm{mm}$ and wider cross-arm $cw = 2 \mathrm{mm}$, confirming a stable polarization rotation condition. In contrast, a narrower cross-arm width $cw = 2 \mathrm{mm}$ leads to a more fluctuating and less predictable phase response, degrading the phase coherence essential for polarization conversion. These results clearly demonstrate that careful tuning of the cross-arm length and width enhances both the magnitude and phase stability of the reflected waves. The observed wideband behavior and consistent phase difference validate the proposed unit cell’s potential for high-performance polarization manipulation in reconfigurable metasurfaces.

To investigate the influence of the stub radius on the electromagnetic performance of the proposed metasurface, Figure 3 illustrate the reflection characteristics and phase behavior. Figure 3a presents the co-polarized reflection coefficients $r_{xx}$ and $r_{yy}$ for various stub radius. It is evident that modifying the radius of the stub significantly affects the fractional bandwidth of the metasurface. Larger stub radius introduce additional resonant modes, which contribute to improved impedance matching and wider operational bandwidths. Figure 3b further supports this observation by showing the phase difference between the cross-polarized reflection coefficients $r_{xy}$ and $r_{yx}$ for different stub radius. As the stub radius increases, the frequency range over which a stable ${180}^{\circ }$ phase difference occurs also expands, which is crucial for achieving effective polarization conversion. A particularly notable case is when the stub radius $rs$ is set to 5 mm, resulting in five distinct resonance points at approximately 3.91 GHz, 4.62 GHz, 7.3 GHz, 12.57 GHz, and 14.84 GHz. These additional matching frequencies arise due to enhanced mutual coupling between adjacent elements, which introduces supplementary resonant modes beyond the fundamental ones. The occurrence of these resonances not only broadens the operating bandwidth but also coincides with the frequencies where a consistent ${180}^{\circ }$ phase difference is observed between $r_{xy}$ and $r_{yx}$, as shown in Figure 3b. This confirms the fulfillment of the polarization conversion condition, specifically when $r_{xx} \approx r_{yy} \approx 0$ and $r_{xy} \approx r_{yx} \approx 1$. Hence, the proposed metasurface achieves efficient and broadband polarization rotation through the careful tuning of the stub radius.

Figure 4 shows the reflection amplitudes of the proposed RRM element under linearly polarized excitation. The co-polarized components ($r_{xx}$ and $r_{yy}$) remain below $-10 \mathrm{dB}$, confirming effective suppression of undesired polarization components. Meanwhile, the cross-polarized reflection coefficients ($r_{xy}$ and $r_{yx}$) exceed $-1 \mathrm{dB}$, indicating strong polarization rotation and efficient conversion across the operating band. This behavior is observed over a wide frequency range from 3.83 GHz to 15.06 GHz, yielding a wide fractional bandwidth of approximately 118.89%—a significant achievement that underscores the metasurface’s ultra-wideband capability. The performance is consistent for both diode states (“ON” and “OFF”), as indicated by the distinct reflection curves and summarized in

Table 1. Bias configurations and corresponding cross-polarized reflection phase.

Diode State	DC Bias	D1	D2	D3	D4	${\mathit{r}}_{\mathbf{xy}}$ Phase
State 1	$-V_f$	on	off	on	off	$0^{\circ }$
State 2	$+V_f$	off	on	off	on	${180}^{\circ }$

. The mirror-symmetric layout of the programmable polarization rotation reflectarray metasurface remains intact across both states, ensuring structural consistency and reliable performance. Figure 5 illustrates the corresponding phase response. A stable phase difference of ${180}^{\circ }$ between the two states is maintained across the entire operating band, confirming reliable 1-bit phase quantization. The coherence between the amplitude and phase responses confirms that the proposed unit cell achieves robust reconfigurability, leveraging the polarization rotation technique to maintain performance over the ultra-wide bandwidth.

3. The 1-Bit RRM Design and Simulation

3.1. RRM Design Configuration

The configuration of the proposed reconfigurable reflectarray metasurface (RRM) is illustrated in Figure 6. It consists of 256 unit cells arranged in a $16 \times 16$ square lattice, specifically designed to achieve dynamic beam steering capabilities. Each unit cell is reconfigurable via a group of PIN diodes, capable of switching between two distinct phase states, thereby supporting binary (1-bit) phase modulation. The total physical aperture of the array is $240 \times 240 {\mathrm{mm}}^2$, corresponding to $8\lambda \times 8\lambda$, where $\lambda$ is the free-space wavelength at the center frequency of 10 GHz. A detailed full-wave simulation model of the entire structure was implemented using the High-Frequency Structure Simulator (HFSS) to analyze and validate the RRM’s far-field performance, as depicted in Figure 7. The reflectarray is illuminated by a y-polarized horn antenna placed at an optimal distance to approximate planar wavefront excitation over the metasurface. This setup enables precise control of the reflected phase distribution by dynamically switching the diode states across the array elements. To achieve effective beamforming, each unit cell is designed to introduce a specific phase compensation such that the reflected wavefronts constructively interfere in the desired direction. The required ideal phase compensation ${\varphi }_{mn}$ for the $(m,n)$-th element, positioned at coordinates ${\overrightarrow{r}}_{mn}$ on the $xy$-plane, is calculated based on the spatial path difference and projected wave vector as follows:

$${\varphi }_{mn} = k_0\left(\left|{\overrightarrow{r}}_{mn} - {\overrightarrow{r}}_f\right| - {\overrightarrow{r}}_{mn}·\overrightarrow{u}\right)$$

(4)

where $k_0 = 2\pi /\lambda$ is the free-space wavenumber, ${\overrightarrow{r}}_f$ is the position vector of the feed antenna, and $\overrightarrow{u}$ is the unit vector indicating the direction of the desired reflected beam. Due to the binary nature of the reconfigurable element, which is realized using a single PIN diode switching between ON and OFF states, only two discrete phase states are supported. To map the continuous phase requirement ${\varphi }_{mn}$ to this 1-bit phase quantization scheme, the calculated phase is first normalized to the interval $(-{90}^{\circ },{270}^{\circ })$ and then quantized using the following rule:

$${\varphi }_{q,mn} = \left\{\begin{array}{cc}{0}^{\circ },\hfill & \mathrm{if} -{90}^{\circ } < {\varphi }_{mn} < {90}^{\circ },\hfill \\ {180}^{\circ },\hfill & \mathrm{otherwise}.\hfill \end{array}\right.$$

(5)

This simple yet effective quantization allows the reflectarray to approximate the desired continuous phase profile with minimal error, thereby enabling real-time beam steering with reduced hardware complexity. The continuous phase pointing in the $({\theta }_{0^{\circ }},{\varphi }_{0^{\circ }})$ beam direction at various scanning angles can be seen in Figure 8. The proposed RRM design demonstrates that even with limited 1-bit phase resolution, efficient far-field beam reconfiguration is achievable across a broad angular range by carefully optimizing the geometry and placement of unit cells.

3.2. Full Wave Simulation and Analysis

In the full-wave simulation model, the RRM and the standard feed horn antenna are placed within two distinct simulation regions, each enclosed using the finite element-boundary integral (FE-BI) method. This hybrid technique significantly reduces computational overhead by limiting the number of mesh elements required in the free-space region between the feed and the RRM, thereby enhancing simulation efficiency without sacrificing accuracy. In addition to improving computational performance, the FE-BI simulation inherently accounts for mutual coupling between adjacent unit cells under plane wave excitation. This ensures that the electromagnetic interactions among elements are fully modeled in the array context. While unit cell behavior was initially characterized under periodic boundary conditions, the full-array simulation confirms stable reflection phase, polarization rotation, and beam steering despite coupling effects. These results validate the robustness of the proposed design in realistic deployment scenarios. The simulated three-dimensional radiation pattern of the proposed RRM at 9.5 GHz is illustrated in Figure 9. The reflectarray achieves a peak gain of 19.9 dBi in the broadside direction (15° reflection), confirming the effective phase control and beamforming capability of the metasurface. This result validates the reconfigurability of the design under ideal excitation and demonstrates the metasurface’s ability to redirect incident energy with high directivity. To further assess the far-field radiation characteristics and mitigate the potential degradation caused by feed horn blockage and resin support interference, the normalized radiation patterns in the $xoz$-plane are examined at multiple frequencies across the operating bandwidth. The dynamic control of phase distributions across the reflectarray aperture enables real-time wide-angle beam steering, reaching up to $\pm {60}^{\circ }$. This wide steering capability highlights the ultra-broadband operational potential of the proposed RRM. Owing to the geometrical symmetry of the reflectarray, bidirectional beam steering is achieved through mirror-symmetric phase arrangements, simplifying the configuration for opposing scan angles. However, due to the dispersive nature of the metasurface elements, the optimal phase coding distribution varies with frequency—even for the same desired beam direction. This frequency-dependent phase response introduces significant complexity when attempting to test across the full bandwidth using individualized coding schemes. To streamline the experimental validation, the broadband frequency range is partitioned into three representative bands centered at 5.5 GHz, 9.5 GHz, and 13.5 GHz. These central frequencies are strategically selected to provide a comprehensive assessment of the RRM’s performance across low, mid, and high-frequency regions of its operational bandwidth. For frequencies near the boundaries between adjacent bands, the coding scheme corresponding to the simulation case with the highest gain is adopted to ensure optimal performance. This strategy effectively reduces the complexity of real-time wideband evaluation while maintaining high fidelity in the beam steering response across multiple frequency bands. Figure 10a,b illustrate the far-field radiation patterns at 5.5 GHz and 13.5 GHz, respectively. Owing to the symmetric structure of the reflectarray, consistent scanning behavior is expected. The reflectarray’s beam-steering capability is analyzed by evaluating scan angles ranging from $0^{\circ }$ to $+ {60}^{\circ }$ in 15° steps. A slight reduction in peak gain is observed when the beam is steered near the broadside direction (0°), primarily attributed to partial blockage from the feed horn. This occurs despite positioning the feed horn at a 17° offset from the z-axis to minimize shadowing effects.

To validate the two-dimensional (2D) angular beamforming capabilities of the proposed RRM, a comprehensive 2D beam scanning analysis was conducted at 9.5 GHz. This analysis covered steering angles up to $\pm {60}^{\circ }$ in both the $xoz$ and $yoz$ principal planes, as illustrated in Figure 10c,d. The simulated results confirmed the formation of stable and well-defined directional beams across the entire scanning range. In the $xoz$-plane, the gain reduction was minimal, with a maximum drop of only 1.52 dB at extreme angles, indicating efficient wide-angle beam steering with negligible performance degradation. Conversely, in the $yoz$-plane, more pronounced gain reductions, reaching up to 3.6 dB, were observed. This asymmetry is primarily attributed to partial wavefront obstruction caused by the physical placement of the feed horn in this plane at higher elevation angles, resulting in feed blockage effects. Despite these angular variances, the RRM consistently produced highly directional pencil beams throughout the operational frequency band from 3.8 GHz to 15 GHz. Figure 11 presents the simulated gain response at broadside ($\theta = 0^{\circ }$), showing a center frequency gain of approximately 19.9 dBi. The peak gain reached 21.3 dBi at 14 GHz, while the minimum gain was 10.6 dBi at 4 GHz. As expected for aperture-based antennas, the gain naturally increased with frequency. Across the wide operating band, the proposed RRM exhibited a gain variation exceeding 3 dB, primarily due to suboptimal coding distributions when operating away from the design frequency. Nevertheless, the gain remained within a 3 dB variation from 8.5 GHz to 15 GHz, resulting in a substantial fractional bandwidth of 55.3%.

To provide a clearer evaluation of the proposed metasurface, several benchmark designs were selected for comparison, as summarized in

Table 2. Comparison between the proposed 1-bit RRM with prior works in size, bandwidth, reconfigurability, and scan range.

Ref.	Element Size $\left({\mathit{\lambda }}^3\right)$	Freq. Range (GHz)	Frac. BW * (%)	Quantization Bit	Scan Range (deg)	Scan Loss (dB)	3 dB Gain BW * (dBi) (%)
[9]	$0.29\times 0.29\times 0.09$	8.3-13	44	1	$\pm {60}^{\circ }$	3.1	32
[10]	$0.22\times 0.22\times 0.07$	6.75–11.25	50	1	$\pm {60}^{\circ }$	NA *	NA *
[11]	$0.26\times 0.26\times 0.06$	8.3–11.5	33	1	$\pm {60}^{\circ }$	3.9	30
[12]	$0.41\times 0.41\times 0.11$	12.9–16.5	24.5	1	$\pm {60}^{\circ }$	4.9	31.5
[24]	$0.43\times 0.43\times 0.05$	11.8–14	17	1	$\pm {50}^{\circ }$	NA*	15.3
[27]	$0.26\times 0.26\times 0.07$	6.82–9.96	37.4	1	${50}^{\circ }$	NA *	NA *
[28]	$0.27\times 0.27\times 0.07$	6.7–10.6	45.1	1	$\pm {60}^{\circ }$	4.8	44.4
This Work	$\mathbf{0}.\mathbf{19}\times \mathbf{0}.\mathbf{19}\times \mathbf{0}.\mathbf{05}$	3.83–15.06	118.89	1	$\pm {\mathbf{60}}^{\circ }$	3.6	55.3

* BW: Bandwidth. NA: Not Available.

. These include recent reconfigurable reflectarray and metasurface structures that offer 1-bit phase modulation over broad frequency ranges, utilizing PIN diode switching mechanisms. The proposed unit cell offers a compact geometric footprint and achieves a significantly wide 3 dB gain bandwidth. While prior works such as [9,10,11,28] demonstrate beam-steering performance, their bandwidths remain comparatively narrow, limiting applicability in broadband systems. In contrast, the proposed RRM, based on a mirror-symmetric configuration, achieves a 3 dB gain bandwidth of 55.3%, effectively spanning 8.5–15 GHz. This improvement enhances its suitability for high-speed, wideband wireless communication applications. For each benchmark, the electrical size of the unit cell, operating frequency band, fractional bandwidth, phase control method, and 2D scanning performance were extracted and tabulated. This enables a comprehensive comparison in terms of reconfigurability, design compactness, and gain stability across scan angles. The proposed design demonstrates the widest bandwidth (118.89%) and supports a broader scan range (up to $\pm {60}^{\circ }$) with only 1-bit control, while also simplifying the structural complexity and biasing requirements.

In summary, the proposed RRM achieves robust two-dimensional beamforming with dynamic phase reconfigurability and wideband polarization conversion. Its performance is characterized by consistent gain levels and beam integrity across both principal scanning planes (xoz and yoz). With these characteristics, combined with its compact form factor and electronic tunability, the design presents a strong candidate for high-performance adaptive beam-steering applications in modern wireless communication environments.

4. Conclusions

In this work, a wideband polarization rotation reflective metasurface (PRRM) with integrated 1-bit reconfigurable phase modulation has been designed and analyzed. The proposed metasurface element utilizes a cross-shaped metallic structure combined with a fan-shaped stub, incorporating vertical interconnect accesses (VIAs) and PIN diodes to achieve dynamic binary control. By employing the mirror-symmetry principle, the structure enables stable 90° polarization conversion along with accurate 1-bit phase modulation across a wide frequency range. Full-wave simulation results confirm that the metasurface achieves cross-polarized reflection coefficients better than –1 dB throughout a broad operating band from 3.83 GHz to 15.06 GHz, corresponding to a relative fractional bandwidth of approximately 118.89%. This performance significantly exceeds the bandwidth offered by several contemporary reconfigurable metasurface designs. Additionally, the reflection magnitude and phase difference remain highly stable across the tuning range, ensuring consistent performance. To demonstrate angular beamforming capabilities, a $16 \times 16$ element array was simulated. The array achieves two-dimensional beam steering with a scanning range of $\pm {60}^{\circ }$ in both principal planes. Notably, the gain response remains stable over a 3 dB gain bandwidth from 8.5 GHz to 15 GHz, resulting in a fractional bandwidth of 55.3%. The design maintains directional pencil beams and minimal gain degradation over the entire angular span. The results confirm that the proposed RRM offers wideband, reconfigurable, and polarization-selective performance with excellent beam steering capabilities. Its compact configuration and broadband behavior make it a strong candidate for next-generation high-speed wireless communication systems and advanced electromagnetic wavefront control applications.