1. Introduction

Due to the advantages of reduced multipath distortion and low polarization deflection loss, high-gain circularly polarized (CP) antennas have been more and more popular in radar, remote sensing and wireless communication systems [1–3]. The conventional method to realize the high-gain property of CP antennas is usually by array antennas or parabolic antennas. However, these two approaches have their own shortcomings. For array antennas, the limited axial ratio (AR) bandwidth and complex feed network usually lead to low radiation efficiencies and can not deal with broadband systems. While the parabolic reflectors to implement phase compensation always require high shape accuracy, which results in high costs, especially for high frequency operation.

As a novel class of antennas with high-gain properties, metasurface (MS)-based reflectarray antennas (RAs) have experienced rapid developments as alternative to conventional array antennas during the past years [4–8]. Owing to their extraordinary capability of tailoring the polarization [9–13], amplitude [14–16] and phase [17–23] of electromagnetic waves, MSs are promising in realizing many exciting functions when applied to RAs [24–27].

Inspired by RAs, transmitarray antennas (TAs) can avoid the issue of feed blockage, and are more popular for their convenient use. Recently, receiver-transmitter (RT) approach has become a common strategy to achieve high transmission efficiency with good AR performance for CP waves [28–33], and several high-gain antennas have been proposed. A linear-to-circular polarization conversion RT MS is designed and applied to a folded TA in [31], and a maximum gain of 22.8 dBi is realized with a 3-dB gain bandwidth of 11.6$\%$. By integrating the RT structures working around 12 GHz and 15 GHz in one meta-atom, a dual-band folded TA antenna working with dual-circular polarization at Ku-band is proposed in [32]. However, aforementioned RT meta-atoms just work in one operation mode, which inevitably results in steep-gain performance and limits their applications in broadband antenna systems.

To alleviate this issue, a novel CPTA enabled by axial-ratio-improved RT MS is proposed in this paper. Split-ring-shaped parasitic patches are introduced to generate another operation mode and improve the AR performance. The transmission magnitude of proposed meta-atoms is higher than −1 dB from 12.5 to 14.5 GHz with the AR lower than 3 dB. The configuration of the proposed CPTA is presented in Fig. 1. A self-made microstrip patch antenna is used as the feed source to generate right-handed circularly polarized (RCP) waves. Combined with TA of phase compensation, a high-gain left-handed circularly polarized (LCP) beam from 12.5 to 15.1 GHz is obtained.

 

Fig. 1. The configuration of the proposed CPTA.

Download Full Size | PDF

2. Design principle of proposed RT-cells

Generally, the electromagnetic waves emitted by CP antennas are elliptically polarized waves in most cases. AR, defined as the ratio of the major axis to the minor axis of the polarization ellipse, is a critical index of CP antennas used to describe the polarization purity. Besides, polarization ratio is commonly applied to calculate AR during actual operation. Denoting $A_{R}(f_{i})$ and $A_{L}(f_{i})$ as the magnitude of RCP component and LCP component in dB at a certain frequency $f_{i}$, respectively, the corresponding $AR(f_{i})$ can be calculated using the following formula:

The structure of traditional patch antennas can be utilized to design RT unit cells [28–33], and probe-fed circular patch loaded with rectangular slot is a classical structure to generate CP operation at a certain frequency $f_{0}$ [34]. However, this type of cells just works in one resonant mode and can achieve a perfect AR within a narrow band. Moreover, the polarization purity deteriorates when the frequency shifts from $f_{0}$, and corresponding AR deteriorates. How to extend the AR bandwidth is an imminently scientific issue. Here, we solve the problem by introducing multi-resonant modes. A split-ring-shaped structure is used to introduce an additional operation mode and extend the working bandwidth of the receiver and transmitter patches. The topology of the proposed broadband RT cells are depicted in Fig. 2. As shown, there are four metal layers. The top receiver patch and one of the metal ground are printed on the first FR4, while the bottom transmitter patch and the other metal ground are printed on the second FR4. These two 2 mm-thick FR4 substrates ($\varepsilon _{r}$ = 2.65 + 0.001i) are bonded together by a 0.1 mm-thick F4B prepreg ($\varepsilon _{r}$ = 3.5 + 0.01i).

 

Fig. 2. The proposed broadband RT meta-atom. (a) Side view, (b) top receiver patch, (c) middle metal ground, (d) bottom transmitter patch, and (e) exploded view. The rotation angle is denoted as $\alpha$, and the incident angle is denoted as $\theta$. The optimized parameters are as follows: p = 10 mm, rp = 2.9 mm, rj = 4.3 mm, ls = 5.4 mm, $\theta _{j}$ = 20 deg, tr = 1.6 mm.

Download Full Size | PDF

To study the operating principle of the proposed receiver/transmitter patches, two probe-fed patches with single resonant mode and multi-resonant modes are discussed. The patches are simulated under periodic boundary conditions. The feed waveguide port and free space port are defined as Port1 and Port2, respectively, as presented in Figs. 3(a) and 3(c). With optimized parameters, patches without parasitic structures (denoted as cell 1) and patches with parasitic structures (denoted as cell 2) both work at 13.5 GHz as displayed in Figs. 3(b) and 3(d). The reflection coefficient $\textrm{S}_{11}$ of cell 1 is lower than −20 dB at 13.5 GHz, and the cell can radiate RCP waves with LCP component lower than −15 dB from 13.1 to 14 GHz. The corresponding AR can be calculated using formula (1). Here, $A_{R}$ and $A_{L}$ can be replaced by $\textrm{S}_{21}$(RCP) and $\textrm{S}_{21}$(LCP), respectively. Obviously, cell 1 just works in one mode with a limited 3-dB AR bandwidth of 6.64$\%$. Loaded with parasitic patches, there are two resonant modes appearing at around 11.3 GHz and 14 GHz, as shown in Fig. 3(d). The LCP component is lower than −15 dB from 12.4 to 14.5 GHz with RCP component higher than −0.5 dB and $\textrm{S}_{11}$ lower than −15 dB. And a 3-dB AR bandwidth of 15.61$\%$ is obtained. Furthermore, the 3-dB AR bandwidth is marked with blue shadow in Figs. 3(b) and 3(d). It can be clearly observed that the AR performance has been enormously improved by introducing parasitic patches.

 

Fig. 3. Operation principle of proposed RT cells. (a) Cell without parasitic patches (cell 1), (b) radiation properties of cell 1, (c) cell with parasitic patches (cell 2), and (d) radiation properties of cell 2.

Download Full Size | PDF

To illustrate the principle mechanism in details, the surface current distribution is depicted in Fig. 4. The surface current of cell 1 at 13.5 GHz mainly concentrates around the end of rectangular slot. Similarly, this can be observed from the first resonant mode of cell 2 at 11.3 GHz as displayed in Fig. 4(b). In this case, there is almost no current on parasitic patches. For the second mode at 14 GHz, current mainly distributes inside the parasitic structure and outside the circular patch. Without doubt, the proposed receiver/transmitter with parasitic patches can work in multi-resonant modes and broaden the working bandwidth.

 

Fig. 4. Surface current distribution of (a) cell 1 at 13 GHz, (b) cell 2 at 11.3 GHz, and (c) cell 2 at 14 GHz.

Download Full Size | PDF

3. Scattering properties of proposed cells

In Section 2, the operation principle of receiver and transmitter is discussed. By cascading transmitter patch and receiver patch with the same handedness, the incident CP waves can be converted to its orthogonal polarization. And by rotating the elements, Pancharatnam-Berry phase can be exploited to offer 2$\pi$ phase modulation [30]. The coupling effect among the adjacent cells is considered, and Floquet ports are adopted to numerically verify the properties of the proposed meta-atoms. The scattering properties of the proposed RT cells after parameter optimization are shown in Fig. 5. The transmission magnitude of cross polarized waves ($\left |t_{LR}\right |$) is as high as −0.8 dB with AR lower than 3 dB when normally illuminated by RCP waves from 12.5 to 14.5 GHz. Corresponding phase shift is controlled by rotation angle which is equal to 2$\alpha$ as presented in Fig. 5(b).

 

Fig. 5. The scattering properties of proposed RT cells under RCP illumination. (a) The magnitude, AR performance and (b) phase shift of transmitted LCP waves varies with rotation angle $\alpha$ when incident angle $\theta$ equals 0 deg. (c) The magnitude and (d) phase shift of transmitted LCP waves varies with incident angle $\theta$ when rotation angle $\alpha$ equals 0 deg. (e) The magnitude and (f) phase shift of transmitted LCP waves varies with rotation angle $\alpha$ when incident angle $\theta$ equals 30 deg.

Download Full Size | PDF

The scattering properties under oblique RCP illumination should also be investigated since the feeding waves are obliquely illuminating the edge of TA. From Figs. 5(c) to 5(f), it can be seen that the transmission magnitude $\left |t_{LR}\right |$ keeps above −1 dB in 12.5-14 GHz even the incident angle is up to 30 deg. Meanwhile, the transmission phase keeps as 2$\alpha$ and remains stable.

4. Design and performance of proposed CPTA

4.1 Design of CPTA

The spherical RCP waves emitted by feed source illustrated in Fig. 1 will be converted to LCP plane waves to improve gain performance via exploiting proposed RT cells as shown in Fig. 2. The compensation phase distribution can be written as:

Herein, (m, n) is the position number of the unit cell, ${\lambda _{0}}$ is the free space wavelength at 13.5 GHz and F is the focal length set as 90 mm. $p_{x}$ and $p_{y}$ are the lengths of unit cell at different directions. To experimentally validate the proposed concept and design, a CPTA consisting of 20 $\times$ 20 cells with a total aperture of 200 $\times$ 200 $mm^{2}$ is designed, fabricated and tested. The phase profile is depicted in Fig. 6(a) and corresponding schematic of the TA is presented in Fig. 6(b). Figure 6(c) shows partial image of the fabricated CPTA and feed antenna. Finally, the CPTA and feed antenna are mounted on an acrylic holder and experimented in the anechoic chamber shown as Fig. 6(d).

 

Fig. 6. The design of proposed CPTA. (a) Phase distribution of CPTA. (b) The schematic of CPTA which consists of 20 $\times$ 20 cells (200 $\times$ 200 $mm^{2}$). (c) Partial image of CPTA and feed antenna. (d) Assembled CPTA in anechoic chamber and measurement setup. The emitted LCP waves propagate along the + z axis, and the radiation patterns in xoz-plane and yoz-plane are measured.

Download Full Size | PDF

4.2 Results and discussion

The S-parameters of feed antenna and CPTA are shown in Fig. 7. The measured impedance bandwidth is smaller than the simulated one, which may be attributed to fabrication errors. The measured −10-dB reflection coefficient bandwidth is 23.5$\%$ covering from 12 to 15.2 GHz. Figure 7(b) compares the gain and AR performance between simulation and measured results. The measured gain shows a 3-dB gain bandwidth of 20.6$\%$ covering from 12.5 to 15.2 GHz and realizes a peak gain of 25.6 dBi at 13.1 GHz, slightly different from the simulation with a 3-dB gain of 21.3$\%$ (from 12.2 to 15.1 GHz) and a peak gain of 25.8 dBi at 13.5 GHz. The measured gain performance deteriorates a little compared with the simulation, and the variations are closely related to fabrication errors, assembly deviation and welding loss. From Fig. 7(b), it can be observed that a 3-dB AR bandwidth of 27.1$\%$ from 12.1 to 15.9 GHz is obtained in the experiment, which is narrower than the simulated one. Besides, Fig. 8 shows the AR performance in xoz-plane and yoz-plane at 13.5 GHz. Within the measured 3-dB beamwidth (around 7$^{\circ }$), both simulated and measured AR keep lower than 3 dB. The performance degradation of AR may be attributed to imperfect receiver CP antenna used in the measurement.

 

Fig. 7. The measured (a) reflection coefficient, (b) gain and AR performance.

Download Full Size | PDF

 

Fig. 8. AR performance in (a) xoz-plane and (b) yoz-plane at 13.5 GHz.

Download Full Size | PDF

The normalized radiation patterns of LCP waves in xoz-plane and yoz-plane of the designed CPTA at 12.5 GHz, 13.5 GHz and 14.5 GHz are presented in Fig. 9. The high-gain pencil beam can be clearly observed, showing that the measured results are in good agreement with the simulated ones. The main properties of the proposed broadband cells and CPTA are also compared with other reported RT MS based antennas as depicted in Table 1. Benefited from the broadband operation of the proposed meta-atoms, although a narrowband CP patch antenna is chosen as the feed source, the proposed transmitarray antenna still achieves a 3-dB gain and AR < 3 bandwidth of 18.8$\%$(from 12.5 to 15.1 GHz), which is better than those in [31–33].

 

Fig. 9. Normalized radiation patterns of LCP waves. (a) xoz-plane at 12.5 GHz. (b) yoz-plane at 12.5 GHz. (c) xoz-plane at 13.5 GHz. (d) yoz-plane at 13.5 GHz. (e) xoz-plane at 14.5 GHz. (f) yoz-plane at 14.5 GHz.

Download Full Size | PDF

Table 1. Comparison between the proposed CPTA and other reported MS-based CP antennas

View Table

5. Conclusion

In conclusion, we have proposed an approach to design broadband CP RT meta-atoms. Numerical simulation shows that parasitic patches can broaden the AR bandwidth by introducing another operation mode. The proposed cells are used to design high-gain CPTA. Both simulated and tested results demonstrate advanced performance of the proposed antenna with an impedance bandwidth of 20.6$\%$ (from 12.5 to 15.2 GHz), a 3-dB AR bandwidth of 27.1$\%$ (form 12.1 to 15.9 GHz) and a 3-dB gain bandwidth of 18.8$\%$ (form 12.5 to 15.1 GHz). The obtained results display that such simply designed, broadband and gain flattened CPTA would be a very attractive candidate for wireless communication systems.