Wideband Polarization Independent Resonant Cavity Antenna for 6G Applications

In this paper, a planar polarization independent resonant cavity antenna (RCA) is proposed to achieve higher gain with a single patch connection instead of using antenna arrays with additional feeding losses. The planar metasurface is based on sub-wavelength complementary circular metallic patches, which provide the same magnitude and phase response for both TE and TM polarized waves. In order to excite the metasurface, an aperture coupled antenna with a circular patch is used to make its integration easier with the metasurface. The proposed antenna design works from 8.97 GHz to 13.1 GHz with a measured 3dB gain bandwidth of 24.2% and it is used to increase the antenna’s gain from 8 dBi up to 14.67 dBi. The RCA achieves the measured bandwidths (input matching ¡-10 dB) above 4 GHz or 37.4%, among the highest reported in the literature with 2 layers.


I. INTRODUCTION
Joint communication and sensing (JC&S) is the central platform for connecting and coordinating billion of devices in the future 6G wireless communication systems [1]. The antenna is the core constituent in the RF front end of these integrated systems and it is accountable for transmitting and receiving electromagnetic waves. The conventional techniques of making an array of an antenna in order to increase the antenna's gain suffer from the substrate and complex feeding network losses [2], [3]. Therefore, there is a growing need to explore alternative and novel approaches to enhance the antenna's gain by minimizing the losses and feeding complexities, especially at the millimeter wave frequencies in order to decrease the path loss and maximize the link margin. For this work, X-band, which can be the future enabler frequency band of 6G applications, is used as the potential frequency band to validate the design.
Metamaterials composed of sub-wavelength resonating unit cell structures with unique properties cover a wide variety of different applications such as high gain resonant cavity antennas (RCA) [2], increasing isolation with electromagnetic bandgaps (EBG) [4], high impedance surfaces (HIS) [5], The associate editor coordinating the review of this manuscript and approving it for publication was Shah Nawaz Burokur . and miniaturized antennas [6]. However, metamaterials suffer from narrowband properties and these properties are polarization dependent [2]. In this paper, we have proposed a wideband polarization independent planar complementary metasurface based on complementary sub-wavelength circular metallic structures showing a constant phase response over a bandwidth of more than 3.7 GHz. The designed metasurface is utilized to enhance the realized gain of the aperture coupled antenna from 8 dBi to maximum 14.67 dBi excluding the feeding network challenges. The metasurface and aperture coupled antenna are fabricated, and integrated together. Finally, the EM simulations and measured results are compared and a close agreement is found between them.
Section II will describe the design, fabrication and measurement results of the resonant cavity antenna (RCA). Finally, the conclusion of the paper is discussed in section III.

II. ANTENNA AND METASURFACE
This section is divided into three subsections. The first subsection gives a detailed overview of the unit cell simulations and extracted magnitude and phase characteristics of the planar metasurface. In the second subsection, the planar metasurface is integrated with the aperture coupled antenna to VOLUME 11, 2023 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ show its focusing characteristics over a wide frequency range.
In the third subsection, the simulation and measured results of the RCA are compared.

A. UNIT CELL SIMULATIONS
The initial phase of evaluating the metasurface is carried out using unit cell simulations while defining the master-slave boundary conditions in the HFSS EM environment as shown in Fig. 1. The planar metasurface is composed of two layers of complementary structures consisting of sub-wavelength symmetrical circular patches. It is designed on a single Rogers 4003C substrate with a thickness of 0.813 mm. Fig. 1 shows the unit cell block of the 2D metasurface with the corresponding highlighted master and slave boundary conditions in the HFSS EM design environment. The unit cell dimensions are 8 mm (Uc x ) x 8 mm (Uc y ) with a top circular patch having a radius of 3.45 mm (Uc R2 ) and the bottom layer consists of an inductive layer made by using circular slots with a radius of 3.15 mm (Uc R1 ) as shown in Fig. 1 (top and bottom sides of the unit cell). The overall summary of the optimized parameters of the unit cell is given in Table 1. The circular structure is chosen for the metasurface because of its symmetry. The unit cell is analyzed with respect to different excited modes (TE and TM polarized waves) by using the Floquet ports in the HFSS EM simulation environment as shown in Fig. 2. The reflection and transmission coefficients of the metasurface with respect to TE and TM polarized waves are shown in Fig. 3. The identical results of the reflection and transmission coefficients with respect to TE and TM polarized waves show that the metasurface is polarization independent when it is excited with respect to different polarized waves. Moreover, the reflection coefficient phase of the metasurface with respect to TE and TM waves is shown in Fig. 3 and the −180 degrees phase requires that the metasurface should be placed λ/2 far away from the H-slot.     Table 1. According to equation 1 [2], the metasurface should be placed λ/2 away from the slot antenna provided that the reflections from the ground are 180 degrees ( g ) and the metasurface introduces a -180 degree ( m ) phase shift (N considered as 1). The estimated theoretical height of the metasurface with respect to different frequencies is shown in Fig. 5, which implies at the lower frequencies, the variation of the height is higher, and this variation becomes less at the higher frequencies. On the other side, the gain of the antenna is dependent on the magnitude of the reflection coefficient as given in equation 2. The theoretically computed directivity with frequency offered by this metasurface is shown in Fig. 5, which indicates an increasing trend of directivity in the frequency regime between 9 GHz to 13 GHz. The total directivity is composed of a summation of the directivity offered by the slot antenna plus the directivity by the metasurface. Considering these variations and the wideband nature of the RCA, the metasurface is placed at a distance of 16 mm above the slot antenna to have a good impedance matching over a bandwidth of around 3.8 GHz. The realized gain also depends highly on this height (h) separation. Lowering it increases the gain at higher frequencies but at the expense of compromising the impedance matching at lower frequencies.
The different factors controlling the input matching of the resonant cavity antenna are also demonstrated in this paper. During these simulations, all the other parameters are set to optimized values, and the influence of the corresponding dimension is recorded on the input reflection of the resonant cavity antenna. The input reflection coefficient response of the resonant cavity antenna (RCA) at different integration levels is shown in Fig. 6. The input reflection coefficient shows two possible resonances without the integration of the patch and the metasurface. As the patch antenna is placed 2 mm above the H-slot, the antenna offers better matching and one extra resonance. Finally, the integration of the top metasurface creates two additional resonances and brings the input reflection below −10 dB covering a frequency bandwidth of 3.8 GHz, as illustrated in Fig. 6.
The influence of the additional stub (with length L stub ) used near the H-slot on the input reflection coefficient of the RCA is shown in Fig. 7. This stub facilitates matching the antenna and its length is chosen large enough, around 1.2 mm to match the integrated antenna over a wide frequency range, especially at the lower frequency band. An investigation was carried out on the H-slot length (L SL ) by varying it from 3.6 mm to 4 mm and its influence was observed on the reflection coefficient of the RCA as shown in Fig. 8. The H-slot length variation has only an impact on the resonance frequency at a lower frequency and the optimized value for this slot length is observed to be 3.8 mm which gives good matching and maximum coverage over the wide frequency range.
Moreover, the influence of the circular patch radius (R p ) is also recorded on the input reflection coefficient of the RCA as depicted in Fig. 9. It can be clearly seen that the circular patch mainly influences the third and last resonance of the RCA. Therefore, its value is chosen to be 5.1 mm to keep the reflection coefficient below −10 dB over a wide frequency range. The height of the metasurface plays a crucial role in deciding its gain in correspondence to good matching. The influence of the metasurface height (h) on the input reflection coefficient of the RCA is demonstrated in Fig. 10. In order to have wideband matching, h is chosen to be 16 mm. Although lower values of h can increase the gain performance of the RCA at higher frequencies, as shown in Fig. 11. However, the matching starts getting worse on the other side.  The metasurface is also analyzed with respect to different polarizations by aligning the slot in horizontal and vertical directions, as shown in Fig. 12. The results are illustrated in Fig. 13, which shows both polarizations have exactly the same behavior on the input reflection coefficient of the antenna. Furthermore, the gain values over the frequency also show that the metasurface increases the gain of the antenna identically, as shown in Fig. 13. Moreover, the polarization independent behavior of the RCA can be realized by rotating the metasurface unit cells with respect to the feeding antenna. Fig. 14 shows the rotation of the circular unit cells has a negligible effect on the reflection coefficient of the antenna. On the other side, the maximum realized gain of the antenna also remains constant considering these unit cells' rotations as shown in Fig. 15.    using plastic spacers at the corner of the boards. Fig. 17 shows the simulated and measured reflection coefficient of   There are a few discrepancies and frequency shifts in the measured results compared to the simulation results, primarily  due to the fabrication tolerance. However, the measured input matching results (S 11 ) are below −10 dB and have a close agreement with the simulation results in the operating frequency range. The measured realized gain of the antenna over frequency is also shown in Fig. 18. This confirms that the metasurface can increase the gain of the antenna up to 14.67 dBi in the regime of 10.56 GHz. The measured gain results (max gain 14.7 dBi) also show the same pattern except for the slight shift in the results owing to the shift in the measured reflection coefficient. The surface current distribution on the top and bottom surfaces of the metasurface is shown in Fig. 19. It shows maximum magnitude at the center and decaying behavior at the edges which comply with the broadside radiation phenomenon.

C. FABRICATION AND MEASUREMENTS
The radiation pattern of the RCA in the E-plane and H-plane is also measured at different frequencies to show its directive behavior. Fig. 20, Fig. 21, and Fig. 22 show the comparison of the RCA's measured and simulated radiation pattern at 9 GHz, 10 GHz and 11 GHz, respectively. There  is a close agreement between the results except for some discrepancies in the E-plane owing to the non-ideal environment and reflections from the positioner and its limited angular resolution in the elevation plane. Table 2 shows a detailed comparison of the proposed work with other recently published research work in terms of different antenna parameters. While comparing with [9], it was found that its 3dB gain bandwidth is a bit higher than our design, but the metasurface presented consists of 4 layers  and it is fed with a waveguide-fed slot which makes it bulky and difficult to integrate for future portable 6G applications. Moreover, its aperture efficiency is less than 35%. Compared with the other two layer designs, it is evident that the metasurface is polarization independent with a 3dB wideband gain and impedance bandwidth characteristics. In addition to that, the maximum radiation efficiency of the antenna is higher than 92% as compared to other reported designs. The    maximum aperture efficiency of the antenna can be further increased by lowering the height of the metasurface but at the expense of decreasing the impedance bandwidth and 3dB  gain bandwidth of the antenna. The maximum aperture efficiency of the resonant cavity antenna can also be increased to 70.7% as given in Table 2 by enclosing it with a 3 mm high vertical metallic shielded wall connected to the ground plane as shown in Fig 23. However, in that case, the 3dB gain bandwidth of the antenna will be decreased, as depicted in Fig. 24 and Fig. 25, respectively.

III. CONCLUSION
This work presents a wideband and high gain resonant cavity antenna that can experimentally work in the frequency range of 8.97 GHz to 13.1 GHz with the fractional impedance bandwidth and 3dB gain bandwidth of 37.4% and 24.2%, respectively. Moreover, given the symmetrical nature of the metasurface, it is shown that the antenna shows identical results if the polarization of the H-slot is changed from horizontal to vertical. As a future work, the tapered version of the metasurface will be analyzed to decrease the sidelobe level further. A special investigation will be done with the tapered surface to keep the gain constant over the 4 GHz frequency bandwidth.