betaFeedback

WorksPhysics and Astronomy

Fixed-Frequency Reconfigurable Leaky-Wave Antennas with Simplified Biasing

Abstract

This work introduces a reconfigurable leaky-waveguide antenna with a simplified biasing scheme for dynamic beam steering at a fixed frequency. Unlike prior metasurface-aided leaky-waveguide antennas that employ tunable metasurfaces as radiative apertures, our approach utilizes them as a waveguide wall solely to control the guided mode. As a result, the proposed structure eliminates the need for local biasing schemes commonly required in earlier designs, significantly simplifying the biasing process. The radiation is achieved by employing passive, angle-independent metasurfaces that allow beam steering across broadside. The feasibility and effectiveness of the proposed design are validated through full-wave simulations.

Introduction

Since their introduction, leaky-waveguide antennas (LWAs) have been a topic of prominent interest due to their number of inherent advantages, such as a low profile, high directivity, and frequency-dependent beam-scanning capabilities [1]. Recently, this interest has been further intensified by advancements in metasurfaces, which have facilitated the development of metasurface-aided LWAs that effectively addressed several longstanding challenges in conventional LWAs. A few notable metasurface-aided LWAs include those utilizing: (i) tunable composite right/left-handed metasurfaces, which overcome the open-stopband effect and enable beam steering through broadside [2, 3]; (ii) switchable metasurfaces with ‘ON’ and ‘OFF’ states, allowing dynamic control of the phase constant of the nthn^{\text{th}} space harmonic coupled to the radiated field [4, 5]; and (iii) Huygens’ metasurfaces, which offer independent control over both the amplitude and phase of the radiated field, allowing for the generation of complex radiation patterns beyond simple beam steering [6, 7].

Regardless of the specific type of metasurfaces employed in metasurface-aided LWAs, their fundamental operating principle remains largely consistent: metasurfaces act as reconfigurable radiative apertures, dynamically converting guided modes into desired radiation. This functionality is made possible by the underlying structure of the metasurfaces, which consist of meta-atoms equipped with active circuit components, such as varactors and PIN diodes. However, this operating principle and design architecture introduce challenges that hinder their practical implementation and scalability. Specifically, steering a beam or shaping the radiation pattern typically requires inhomogeneous surface properties [8]. As such, the local interaction between a guided mode and meta-atoms has to be precisely controlled, which in turn necessitates individual biasing of each meta-atom (i.e., a local biasing). Therefore, complex and cumbersome biasing networks are needed that become increasingly difficult to manage as the systems scale.

To address this limitation, we propose a metasurface-aided LWA that achieves dynamic beam steering at a fixed operating frequency while significantly simplifying the biasing scheme. Unlike previous designs that employ tunable metasurfaces as radiative apertures, our tunable metasurface does not radiate any fields but only modulate the guided mode within the LWA. This key distinction allows the proposed design to operate with global biasing (i.e., uniform biasing), entirely eliminating the need for the local biasing schemes required in earlier metasurface-aided LWAs. Meanwhile, the radiative aperture in our design remains entirely passive. Specifically, it consists of a pair of passive angle-independent metasurfaces that directly transform the guided mode into a highly directive beam, with its direction controlled by adjusting the propagation angle of the guided mode. The proposed metasurface-aided LWA is well-suited for applications where ease of fabrication and simplified biasing are critical, ranging from wireless communication networks to sensing and imaging systems.

Theory and Concept

Figure 1 illustrates the schematic of the proposed metasurface-aided reconfigurable LWA, which consists of three main components: (i) a tunable reflective metasurface (TRM), (ii) an angle-independent partially reflective surface (PRS), and (iii) an angle-independent binary metasurface (BMS). Before detailing the underlying principle of the proposed LWA, it is important to note that both the TRM and PRS are homogeneous surfaces, meaning their surface properties remain uniform across their entire area. For the TRM, this homogeneity physically implies that its unit cells are connected in parallel and uniformly biased. This uniform biasing can be easily implemented using global bias lines along the waveguide direction, yy, as depicted in the inset of Figure 1(a). On the other hand, the PRS is modeled as a cascade of three admittance surfaces (YtopY_{\text{top}}, YmidY_{\text{mid}}, and YbotY_{\text{bot}}), separated by a 1.575-mm thick Rogers 5880 substrate (εr=2.2\varepsilon_r = 2.2, δ=0.0009\delta= 0.0009). Each of these admittance surfaces is assumed to comprise entirely passive and identical unit cells to ensure homogeneity.

A schematic of the proposed reconfigurable LWA with simplified biasing

Figure 1. A schematic of the proposed reconfigurable LWA with simplified biasing. (a) Overview of the proposed reconfigurable LWA design comprising tunable reflective metasurface (TRM), angle-independent partially-reflective surface (PRS), and angle-independent binary metasurface. (b) Cross-section of the proposed LWA.

Hereafter, we denote the reflection coefficients of TRM and PRS as ΓTRM\Gamma_{\text{TRM}} and ΓPRS\Gamma_{\text{PRS}}, respectively. A guided mode can then be formed between the TRM and PRS, provided that the guidance resonance condition is satisfied, as

∠ΓTRM+∠ΓPRS−2k0dcos⁡(θin)=0(1)\angle\Gamma_{\text{TRM}}+\angle\Gamma_{\text{PRS}}-2k_0d\cos(\theta_{\text{in}})=0 \tag*{(1)}

where θin\theta_{\text{in}}, k0k_0, and dd represents the propagation angle of the guided mode, the wavenumber in free space, and the distance between the TRM and PRS, respectively (see Figure 1(b)).

θin\theta_{\mathrm{in}}20∘20^\circ30∘30^\circ40∘40^\circ50∘50^\circ60∘60^\circ70∘70^\circ
∠ΓTRM\angle\Gamma_{\mathrm{TRM}}−66.4∘-66.4^\circ−104∘-104^\circ−154∘-154^\circ144∘144^\circ72∘72^\circ−7.6∘-7.6^\circ

Table 1. Summary of the required reflective phase with respect to the propagation angle of the guided mode.

From Eq. (1), it is seen that θin\theta_{\mathrm{in}} can be dynamically tuned as a function of ∠ΓTRM\angle\Gamma_{\mathrm{TRM}}, and Table I summarizes the required phase of ΓTRM\Gamma_{\mathrm{TRM}} for synthesizing various θin\theta_{\mathrm{in}} when d=0.5λod = 0.5\lambda_{\mathrm{o}}. It is noted that variations in the bias voltage induce adjustments in the reflection phase of TRM, which subsequently modifies the propagation angle in the LWA. Furthermore, if the magnitude of ΓPRS\Gamma_{\mathrm{PRS}} is kept below unity, the PRS would radiate a beam at θin\theta_{\mathrm{in}} (i.e., the fundamental-mode radiation), which varies as ∠ΓTRM\angle\Gamma_{\mathrm{TRM}} changes according to Eq. (1). Thus, the combination of the TRM and PRS facilitates beam steering with a simple biasing mechanism. However, the reflection properties of the PRS are not guaranteed to remain independent of θin\theta_{\mathrm{in}}. Moreover, this method fails to achieve broadside beam steering as the fundamental mode is radiated.

To first address the angular dependence of the PRS, ΓPRS\Gamma_{\mathrm{PRS}} is evaluated as a function of the propagation angle, θin\theta_{\mathrm{in}}, using the transverse equivalent circuit model (see the inset of Figure 1(a)), expressed as [9],

ΓPRS=ρ0+ρ1kz,oko−1+ρ2(kz,oko−1)2δ0+δ1kz,oko−1+δ2(kz,oko−1)2+δ3(kz,oko−1)3(2)\Gamma_{\mathrm{PRS}} = \frac{\rho_{0}+\rho_{1}k_{z,\mathrm{o}}k_{\mathrm{o}}^{-1}+\rho_{2}(k_{z,\mathrm{o}}k_{\mathrm{o}}^{-1})^{2}}{\delta_{0}+\delta_{1}k_{z,\mathrm{o}}k_{\mathrm{o}}^{-1}+\delta_{2}(k_{z,\mathrm{o}}k_{\mathrm{o}}^{-1})^{2}+\delta_{3}(k_{z,\mathrm{o}}k_{\mathrm{o}}^{-1})^{3}} \tag*{(2)}

where kz,o=kocos⁡(θin)k_{z,\mathrm{o}} = k_{\mathrm{o}}\cos(\theta_{\mathrm{in}}) and the remaining coefficients are defined as [9],

ρ0=jq[ξbot(ξtopξmid−2ξ)−ξ(ξtop−ξbot)],(3)\rho_{0} = \frac{j}{q}\left[\xi_{\mathrm{bot}}(\xi_{\mathrm{top}}\xi_{\mathrm{mid}}-2\xi)-\xi(\xi_{\mathrm{top}}-\xi_{\mathrm{bot}})\right], \tag*{(3)}
ρ1=ξmid(ξtop−ξbot),(4)\rho_{1} = \xi_{\mathrm{mid}}(\xi_{\mathrm{top}}-\xi_{\mathrm{bot}}), \tag*{(4)}
ρ2=jqξmid−ju(ξtop+ξbot),(5)\rho_{2} = jq\xi_{\mathrm{mid}}-ju(\xi_{\mathrm{top}}+\xi_{\mathrm{bot}}), \tag*{(5)}
δ0=−ρ0,(6)\delta_{0} = -\rho_{0}, \tag*{(6)}
δ1=ξmid(ξtop+ξbot)−2ξ,(7)\delta_{1} = \xi_{\mathrm{mid}}(\xi_{\mathrm{top}}+\xi_{\mathrm{bot}})-2\xi, \tag*{(7)}
δ2=jqξmid+ju(ξtop+ξbot),(8)\delta_{2} = jq\xi_{\mathrm{mid}}+ju(\xi_{\mathrm{top}}+\xi_{\mathrm{bot}}), \tag*{(8)}
δ3=−2qu(9)\delta_{3} = -2qu \tag*{(9)}

Here, ξtop\xi_{\mathrm{top}}, ξmid\xi_{\mathrm{mid}}, and ξbot\xi_{\mathrm{bot}} represent the normalized admittance values in each admittance surfaces forming the PRS. Particularly, they are defined as

ξtop=1+jqYtopηo,(10)\xi_{\mathrm{top}} = 1 + jqY_{\mathrm{top}}\eta_{\mathrm{o}}, \tag*{(10)}
ξmid=2+jqYmidηo,(11)\xi_{\mathrm{mid}} = 2 + jqY_{\mathrm{mid}}\eta_{\mathrm{o}}, \tag*{(11)}
ξbot=1+jqYbotηo,(12)\xi_{\mathrm{bot}} = 1 + jqY_{\mathrm{bot}}\eta_{\mathrm{o}}, \tag*{(12)}

where ξ=1+pq\xi= 1 + pq. Based on Eq. (2), the values of ξtop\xi_{\mathrm{top}}, ξmid\xi_{\mathrm{mid}}, and ξbot\xi_{\mathrm{bot}} are optimized to achieve an angle-independent reflection coefficient. Specifically, a particle swarm optimization is used to optimize ξtop\xi_{\mathrm{top}}, ξmid\xi_{\mathrm{mid}}, and ξbot\xi_{\mathrm{bot}} such that the magnitude of ΓPRS\Gamma_{\mathrm{PRS}} is fixed at 0.70.7 for θin∈[20∘,70∘]\theta_{\mathrm{in}} \in[20^\circ,70^\circ]. The optimized admittance surface values are Ytop,PRS=−0.0033Ω−1Y_{\mathrm{top,PRS}}=-0.0033\Omega^{-1}, Ymid,PRS=−0.0059Ω−1Y_{\mathrm{mid,PRS}}=-0.0059\Omega^{-1}, and Ybot,PRS=−0.0179Ω−1Y_{\mathrm{bot,PRS}}=-0.0179\Omega^{-1}. Figure 2 shows the resulting reflection coefficient as a function of θin\theta_{\mathrm{in}}, obtained using Floquet mode simulations in Ansys HFSS. As shown, the magnitude of ΓPRS\Gamma_{\mathrm{PRS}} remains close to the target value of 0.70.7, while the phase remains relatively constant over the specified range of angles.

Plot of the Floquet mode simulation results depicting the transmittance characteristics of the angle-independent PRS

Figure 2. Plot of the Floquet mode simulation results depicting the transmittance characteristics of the angle-independent PRS.

In addition, to mitigate the open-stopband effect, we invoke the theory of Fourier optics which states that an aperture field and the corresponding Fraunhofer radiation form a Fourier transform pair. In light of this theory, a passive BMS is placed on top of the PRS (see Figure 1) to convert the fundamental-mode radiation from the PRS into the nthn^{\mathrm{th}} spatial harmonic. For this purpose, two types of unit cells are arranged to provide fixed transmission phases of +160∘+160^\circ and −20∘-20^\circ, with near-unity transmission magnitude, thereby creating a binary transmission profile expressed as

S21,bin=sgn⁡(cos⁡(2πyP))e−j20∘,(13)S_{21,\mathrm{bin}} = \operatorname{sgn}\left(\cos\left(\frac{2\pi y}{P}\right)\right)e^{-j20^\circ}, \tag*{(13)}

where sgn denotes the signum function, and PP represents the periodicity of the binary pattern, defined as [6],

P=2π∣kosin⁡θa−kosin⁡θb∣.(14)P = \frac{2\pi}{\left|k_{\mathrm{o}}\sin\theta_{a}-k_{\mathrm{o}}\sin\theta_{b}\right|}. \tag*{(14)}

Here, θa\theta_{a} and θb\theta_{b} are set to 50∘50^\circ and 0∘0^\circ, respectively. These angles are chosen to ensure that only one Fourier component, which corresponds to one propagating mode, resides within the visible region when the propagation angle of the guided mode varies from 20∘20^\circ to 70∘70^\circ. In this way, when the radiation angle (i.e. θin\theta_{\mathrm{in}}) from the PRS changes due to the TRM around 50∘50^\circ, the output angle from the BMS can also change around broadside, thereby overcoming the open-stopband effect.

Finally, to ensure that the binary transmission profile in Eq. (13) also remains consistent across all the incident angles within {20∘,70∘}\{20^\circ, 70^\circ\}, the angular dependence of the transmission coefficient, TBMST_{\mathrm{BMS}}, is evaluated as [9],

TBMS=τ1kz,oko−1+τ3(kz,oko−1)3δ0+δ1(kz,oko−1)+δ2(kz,oko−1)2+δ3(kz,oko−1)3(15)T_{\mathrm{BMS}} = \frac{\tau_1 k_{z,o}k_o^{-1}+\tau_3\left(k_{z,o}k_o^{-1}\right)^3}{\delta_0+\delta_1\left(k_{z,o}k_o^{-1}\right)+\delta_2\left(k_{z,o}k_o^{-1}\right)^2+\delta_3\left(k_{z,o}k_o^{-1}\right)^3} \tag*{(15)}

where τ1=2ξ\tau_1=2\xi and τ3=2q(2u−q)\tau_3=2q(2u-q). Based on Eq. (15), the admittance surfaces are optimized for full transmission, with the transmission phase set to either 160∘160^\circ and −20∘-20^\circ for θin∈{20∘,70∘}\theta_{\mathrm{in}}\in\{20^\circ,70^\circ\}. For a transmission phase of 160∘160^\circ, the corresponding admittance surfaces are calculated as Ytop,+160=−0.0079 Ω−1Y_{\mathrm{top},+160}=-0.0079\ \Omega^{-1}, Ymid,+160=−0.0323 Ω−1Y_{\mathrm{mid},+160}=-0.0323\ \Omega^{-1}, and Ybot,+160=−0.0079 Ω−1Y_{\mathrm{bot},+160}=-0.0079\ \Omega^{-1}. On the other hand, for a transmission phase of −20∘-20^\circ, the calculated admittance surfaces are Ytop,−20=−0.001 Ω−1Y_{\mathrm{top},-20}=-0.001\ \Omega^{-1}, Ymid,−20=0.0048 Ω−1Y_{\mathrm{mid},-20}=0.0048\ \Omega^{-1}, and Ybot,−20=−0.001 Ω−1Y_{\mathrm{bot},-20}=-0.001\ \Omega^{-1}. It is noted that, similar to before, these two unit cells are modeled as a cascade of three admittance surfaces, separated by a 1.575-mm thick Rogers 5880 substrate. Figure 3 plots the optimized magnitude and phase variation with respect to the incident angles. As seen, the magnitude of TBMST_{\mathrm{BMS}} is observed to approach near unity, and the phase also remains relatively constant. These findings suggest promising alignment with theoretical expectations, as will become evident through full-wave demonstrations presented in the following section.

Comparison of the transmittance of the BMS as obtained from HFSS simulations and MATLAB calculations

Figure 3. Comparison of the transmittance of the BMS as obtained from HFSS simulations and MATLAB calculations. (a) Magnitude of the 160∘160^\circ BMS, (b) Phase of the 160∘160^\circ BMS, (c) Magnitude of the −20∘-20^\circ BMS, and (d) Phase of the −20∘-20^\circ BMS.

Full-wave demonstration

To validate the proposed concept, an initial investigation is conducted by simulating the LWA depicted in Figure 1 in the absence of the BMS via ANSYS HFSS. In this simulation, the waveguide length is set to 10λo10\lambda_o, and the TRM and PRS are modeled as the impedance boundaries to minimize computational costs. Without the BMS, it is expected that the PRS radiates the fundamental mode that propagates at θin\theta_{\mathrm{in}}. By applying a bias voltage across the TRM, the reflective phase can be adjusted and alter θin\theta_{\mathrm{in}}. Consequently, although our radiative aperture (i.e. the PRS) is entirely passive, the radiated beam can be dynamically steered. This is illustrated in Figure 4 from which it is seen that the fundamental mode

Full-wave radiation patterns (gains) at various radiation angles for the fundamental mode radiation

Figure 4. Full-wave radiation patterns (gains) at various radiation angles for the fundamental mode radiation.

radiates at various θin\theta_{\mathrm{in}}, which aligns well with the propagation angle within the LWA. It should be noted that Ohadi et al. have recently proposed a tunable LWA that also allows dynamic steering of the fundamental mode by varying the propagation angle of a guided mode [10]. In their work, however, two tunable metasurfaces were required, while the proposed work only utilizes one surface, thereby further simplifying the design architecture and biasing.

Over the beam-steering range of 20∘20^\circ to 70∘70^\circ, the gain experiences a degradation of approximately 4 dB. This drop would be substantially larger if the PRS were not optimized for angle-independence, as non-optimized surfaces typically fail to maintain a consistent transmissive magnitude at oblique incidences.

Although Figure 4 demonstrates that our simple biasing scheme enables beam steering, it does not address the open-stopband effect because the fundamental mode is still being radiated. To resolve the open-stopband problem, we now consider the case where the BMS is incorporated, with its transmission profile defined in Eq. (13). Specifically, the BMS is placed 1 mm above the PRS, and a secondary excitation port is introduced at the opposite end of the waveguide to extend the range of propagation angle from −70∘-70^\circ to +70∘+70^\circ. It is reminded that the period of the binary pattern is chosen such that the LWA would radiate at broadside when the propagation angle corresponds to 50∘50^\circ. Figure 5 shows the full-wave simulation results of the proposed LWA with the BMS. As shown, when the propagation angle corresponds to 50∘50^\circ, the maximum radiation occurs at θout=1∘\theta_{\mathrm{out}} = 1^\circ. Furthermore, as the propagation angle changes around 50∘50^\circ, the radiation angle correspondingly scans across broadside, effectively eliminating the open-stopband effect.

Full-wave radiation patterns (gains) at various radiation angles

Figure 5. Full-wave radiation patterns (gains) at various radiation angles. In this context, “port 1” denotes the scenario in which only port 1 is excited, with port 2 assumed to be perfectly matched. Conversely, “port 2” represents the complementary scenario, where only port 2 is excited under the same assumption.

Conclusion

This work presents a route to achieving dynamic beam steering at a fixed frequency using a novel metasurface-aided leaky-waveguide antenna that allows simplified biasing scheme. The design incorporates a tunable reflective metasurface (TRM) within the waveguide, which dynamically controls the propagation angle of a guided mode by varying its reflective phase. Simplified biasing is achieved by connecting all the unit cells of the TRM in parallel. For radiation, the partially reflective surface (PRS) is designed with partial transmission and angle-independent characteristics, ensuring efficient and consistent fundamental-mode radiation. Additionally, the binary metasurface (BMS) positioned above the PRS addresses the open-stopband problem. The feasibility of the proposed concept has been demonstrated through full-wave simulations using Ansys HFSS. Future efforts will focus on the physical realization of the proposed LWA, including experimental validation and further optimization for practical applications in dynamic beamforming technologies.

Acknowledgments

This work was supported by the National Research Foundation of Korea (NRF) grants funded by the Korea government (MSIT) (RS-2024-00341191 and RS-2024-00343372).

References

  1. [1]D. K. Karmokar, Y. J. Guo, S.-L. Chen, and T. S. Bird, “Composite right/left-handed leaky-wave antennas for wide-angle beam scanning with flexibly chosen frequency range,” IEEE Transactions on Antennas and Propagation, vol. 68, no. 1, pp. 100–110, 2020.DOI
  2. [2]C. Caloz, T. Itoh, and A. Rennings, “CRLH metamaterial leaky-wave and resonant antennas,” IEEE Antennas and Propagation Magazine, vol. 50, no. 5, pp. 25–39, 2008.DOI
  3. [3]C. Damm, M. Maasch, R. Gonzalo, and R. Jakoby, “Tunable composite right/left-handed leaky wave antenna based on a rectangular waveguide using liquid crystals,” in 2010 IEEE MTT-S International Microwave Symposium, pp. 13–16, 2010.
  4. [4]D. K. Karmokar, D. N. P. Thal akotuna, K. P. Esselle, M. Heimlich, and L. Matekovits, “Fixed-frequency beam steering from a stub-loaded microstrip leaky-wave antenna,” in 2013 International Symposium on Electromagnetic Theory, pp. 436–439, 2013.
  5. [5]Z. Li, Y. J. Guo, S.-L. Chen, and J. Wang, “A period-reconfigurable leaky-wave antenna with fixed-frequency and wide-angle beam scanning,” IEEE Transactions on Antennas and Propagation, vol. 67, no. 6, pp. 3720–3732, 2019.DOI
  6. [6]E. Abdo-Sánchez, M. Chen, A. Epstein, and G. V. Eleftheriades, “A leaky-wave antenna with controlled radiation using a bianisotropic Huygens’ metasurface,” IEEE Transactions on Antennas and Propagation, vol. 67, no. 1, pp. 108–120, 2018.
  7. [7]M. Kim and G. V. Eleftheriades, “Guided-wave-excited binary Huygens’ metasurfaces for dynamic radiated-beam shaping with independent gain and scan-angle control,” Physical Review Applied, vol. 15, no. 5, p. 054037, 2021.DOI
  8. [8]M. Imbert, A. Papío, F. De Flaviis, L. Jofre, and J. Romeu, “Design and performance evaluation of a dielectric flat lens antenna for millimeter-wave applications,” IEEE Antennas and Wireless Propagation Letters, vol. 14, pp. 342–345, 2015.
  9. [9]A. Shaham and A. M. Epstein, “All-angle nonlocal metasurfaces on demand: Universal realization of normal susceptibilities via multilayered printed-circuit-board (PCB) cascades,” arXiv preprint arXiv:2412.17173, 2024.arxiv.org/abs/2412.17173
  10. [10]A. Ohadi and G. V. Eleftheriades, “Fixed-frequency beam-steering using slotted waveguide with tunable impedance walls,” IEEE Open Journal of Antennas and Propagation, vol. 2, pp. 978–990, 2021.

Paper details

vlnc:W_8e2gacjb
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents