A Novel Design and Development of a Strip-Fed Circularly Polarized Rectangular Dielectric Resonator Antenna for 5G NR Sub-6 GHz Band Applications

In this article, a rectangular dielectric resonator antenna (RDRA) with circularly polarized (CP) response is presented for 5G NR (New Radio) Sub-6 GHz band applications. A uniquely shaped conformal metal feeding strip is proposed to excite the RDRA in higher-order mode for high gain utilization. By using the proposed feeding mechanism, the degenerate mode pair of the first higher-order, i.e., TEδ13x at 4.13 GHz and TE1δ3y, at 4.52 GHz is excited to achieve a circularly polarized response. A circular polarization over a bandwidth of ~10%, in conjunction with a wide impedance matching over a bandwidth of ~17%, were attained by the antenna. The CP antenna proposed offers a useful gain of ~6.2 dBic. The achieved CP bandwidth of the RDRA is good enough to cover the targeted 5G NR bands around 4.4–4.8 GHz, such as n79. The proposed antenna configuration is modelled and optimized using computer simulation technology (CST). A prototype was built to confirm (validate) the performance estimated through simulation. A good agreement was observed between simulated and measured results.


Introduction
The 5G New Radio (NR) is a newly developed air interface to fulfill the requirements of modern communications. 5G wireless technology is becoming popular because of its significant features, such as high data rate, low latency response time, and high bandwidth [1]. Figure 1 presents the configuration of the proposed geometry of the circularly polarized antenna excited by a uniquely shaped conformal strip placed with a feed point at the central position. The design was modeled in CST ® Microwave Studio using its time domain-based finite integration technique [23]. Hexahedral meshing was used to design a rectangular DRA with DR permittivity, ε r = 10 [11]. In settings, the cell per wavelength was set at a value of 40 and the cells per max model box edge at 20 for design mashing. The fraction of maximum cell near to model was also set at a value of 20 and total unknowns of 559,980 were obtained.

Antenna Geometry
3, antenna design, optimization, and CP generation mechanism are discussed in d Section 4 comparison between experiment and theory is demonstrated and explain tion 5 features the conclusion to the article. Figure 1 presents the configuration of the proposed geometry of the circularl ized antenna excited by a uniquely shaped conformal strip placed with a feed poi central position. The design was modeled in CST ® Microwave Studio using its t main-based finite integration technique [23]. Hexahedral meshing was used to d rectangular DRA with DR permittivity, ε = 10 [11]. In settings, the cell per wav was set at a value of 40 and the cells per max model box edge at 20 for design m The fraction of maximum cell near to model was also set at a value of 20 and t knowns of 559,980 were obtained. An iterative design procedure was followed to determine the optimum dim of the feeding metallic strips that were needed to excite the degenerate TE an necessary for CP wave generation [24]. The DR profile dimensions of H = 26.1 m 25.4 mm, and D = 14.3 mm were used. The optimization of the lengths and the w the strip were performed by running different simulations using many parameter The results of the design procedure are summarized in Table 1, which shows sev lected dimensions of l1, l2, l3, l4, and l5 that could generate circular polarization in c tion with sufficient impedance matching bandwidth.  An iterative design procedure was followed to determine the optimum dimensions of the feeding metallic strips that were needed to excite the degenerate TE x δ13 and TE y 1δ3 necessary for CP wave generation [24]. The DR profile dimensions of H = 26.1 mm, W = 25.4 mm, and D = 14.3 mm were used. The optimization of the lengths and the widths of the strip were performed by running different simulations using many parameter sweeps. The results of the design procedure are summarized in Table 1, which shows several selected dimensions of l 1 , l 2 , l 3 , l 4 , and l 5 that could generate circular polarization in conjunction with sufficient impedance matching bandwidth.

Circularly Polarized 5G Antenna Design and Optimization
In this section, the design and development of CP 5G NR Sub-6 GHz RDRA are explained in detail. Moreover, the antenna optimization, along with results, is discussed. The lengths and widths of the feeding strip were very critical in this design configuration which is herein demonstrated and discussed. It is a well-known fact that a circularly polarized wave can be generated if the feeding network is capable of exciting the orthogonal degenerate modes, since such excitation generates two far-field components that are equal in magnitude with the quadrature phase shift necessary for CP generation [25]. In this section an initially linearly polarized rectangular DRA was designed using a single conformal strip. The next feeding mechanism was modified step by step to design a circularly polarized DRA without adding any complicated changes. Five different design geometries for the development of CP 5G Sub-6 GHz RDRA are presented and discussed in detail.

Geometry 1
In Figure 2a geometry 1 of RDRA is depicted. As shown, the antenna was designed by placing a single conformal strip at the middle of the surface of the DRA. The antenna was linearly polarized because such excitation does not excite the degenerate mode. Only TE y 1δ3 was energized, at around 4.47 GHz, to generate a linearly polarized wave, as shown in Figure 3. The result of the axial ratio of geometry 1 is not shown in Figure 4.

Geometry 2
The geometry 2 of RDRA is shown in Figure 2b. The feeding strip was modified by adding one more length. Using this configuration, the RDRA showed behavior towards CP response. The degenerate mode was excited but still, S11 bandwidth below −10 dB was nil, as shown in Figure 3. As presented in Figure 3, the degenerate mode pair of the first

Geometry 2
The geometry 2 of RDRA is shown in Figure 2b. The feeding strip was modified by adding one more length. Using this configuration, the RDRA showed behavior towards CP response. The degenerate mode was excited but still, S11 bandwidth below −10 dB was nil, as shown in Figure 3. As presented in Figure 3, the degenerate mode pair of the first

Geometry 2
The geometry 2 of RDRA is shown in Figure 2b. The feeding strip was modified by adding one more length. Using this configuration, the RDRA showed behavior towards CP response. The degenerate mode was excited but still, S 11 bandwidth below −10 dB was nil, as shown in Figure 3. As presented in Figure 3, the degenerate mode pair of the first higher-order, i.e., TE x δ13 and TE y 1δ3 , was excited at 4.12 GHz and 4.66 GHz, respectively. The CP response of the antenna is depicted in Figure 4, and, as shown, the 3-dB axial ratio bandwidth was nil as well. The results of this configuration were encouraging, but needed further modification and optimization to achieve the desired bandwidths for the targeted application. Geometry 2 was further modified, and the impact of the modification is discussed in the next section.

Geometry 3
In geometry 3 the RDRA was energized by adding one more strip, as presented in Figure 2c. The modification is discussed in the next section. The return losses of this design configuration are depicted in Figure 3. The S 11 curve of geometry 3 shows that the degenerate TE x δ13 was excited at 4.16 GHz and TE y 1δ3 at 4.4 GHz, but still, impedance matching (|S 11 | ≤ 10 dB) was not achieved. As shown in Figure 4, the 3 dB axial ratio bandwidth of~2% was achieved. The provided bandwidth was not enough to cover the targeted 5G band. Moreover, the S 11 of the antenna needed to be optimized further. So, the shape of the feeding strip was further changed and this is discussed in the next geometry.

Geometry 4
The geometry 4 of RDRA is presented in Figure 2d. This design configuration was achieved by adding an additional length to the feeding strip. At this stage, the feeding strip was composed of 4 different lengths. The impedance matching (|S 11 | ≤ 10 dB) over a bandwidth of~8.8% was attained by this configuration, as shown in Figure 3. The degenerate mode pair of the first higher-order, i.e., TE x δ13 at 4.14 GHz and TE y 1δ3 , at 4.55 GHz was excited to generate the circularly polarized response. The circular polarization over a bandwidth of~5.8% was provided by the antenna. A significant improvement was observed in S 11 and AR ratio bandwidths but, still, the achieved CP response was not enough to cover the desired bandwidth i.e., 4.4-4.8 GHz. This design configuration was again modified to make the final proposed geometry.

Geometry of the Proposed CP 5G RDRA
The final design geometry was developed by adding the fifth and last length to the feeding strip to make the desired 5G Sub-6 GHz rectangular DRA, as depicted in Figures 1 and 2b. The optimized strip lengths were l 1 = 5.5 mm, l 2 = 7 mm, l 3 = 5 mm, l 4 = 10 mm, and l 5 = 1.5 mm. The feed parameters were optimized by running a number of simulation sweeps to get the desired CP response. An impedance matching (|S 11 | ≤ 10 dB) over a bandwidth of~17% was provided by the proposed geometry, as presented in Figure 3. As shown, the degenerate modes TE x δ13 at 4.13 GHz and TE y 1δ3 at 4.52 GHz were excited to generate the circularly polarized response. The E-field and H-field distribution of the proposed CP 5G RDRA aare depicted in Figures 5 and 6, respectively.
The CP response over a bandwidth of~10% was provided by the antenna. The achieved 3 dB axial ratio extended from 4.4-4.84 GHz which was good enough to cover the targeted n79 band. The AR and S11 bandwidths were achieved over the same range, as shown in Figure 7. Moreover, as the DRA radiated away from the ground the size did affect the performance much [26]. The antenna was simulated on different ground plane sizes and the results of S11 and the axial ratio are presented in Figures 8 and 9. The simulated surface current distributions of the antenna at 4.13 GHz (Minimum of S11) are depicted in Figure 10. As can be seen the composite current surface currents on the novel feed were orthogonal at 0 • and 90 • , which provided the required condition for CP generation.            The wideband CP response and higher-order mode excitation were achieved using a low-cost simple design configuration without any complexity, which is a good contribution to those reported in the literature. The performance comparison of different geometries in the development of the desired CP 5G Sub-6 GHz antenna is summarized in Table 2. The optimized circularly polarized antenna was then fabricated to experimentally validate the proposed design which is demonstrated and explained in the next section.

Measurement Results
The optimized 5G NR band circularly polarized antenna was finally fabricated measure the experimental results. The photographs of the proposed prototype presented in Figure 11. The closeup, front view, top view, and back view of the ante are depicted in Figure 11a-d, respectively. The ECCOSTOCK HiK with permittivity, 10 having loss tangent (δ) of 0.002 was used as DR material. An 80 × 80 mm alumin ground plane was used. The feeding strip was cut from adhesive copper tape to ea stick to the surface of the DRA. An SMA was soldered at the feed point at the center of DRA with the feeding strip. S11 was measured using a vector network analyzer (VN while far-field parameters were measured in the anechoic chamber. A 50-Ω coaxial ca was used to connect the SMA with the VNA. Double-sided copper tape was used to s the DRA on the aluminum ground plane] to remove the possible air gap, according to procedure explained in [26].
The resonant mode frequencies of the RDRA could be predicted using mathemat equations of the dielectric waveguide model, as explained in [27]. According to DW TE was estimated at 3.89 GHz and TE at 4.53 GHz. The comparison betw simulated and measured S11 of the CP DRA is presented in Figure 12. TE was simula at 4.13 GHz and measured at 4.12 GHz while TE was simulated at 4.52 GHz  in Table 2. The optimized circularly polarized antenna was then fabricated to experimentally validate the proposed design which is demonstrated and explained in the next section. in Table 2. The optimized circularly polarized antenna was then fabricated to experimentally validate the proposed design which is demonstrated and explained in the next section. in Table 2. The optimized circularly polarized antenna was then fabricated to experimentally validate the proposed design which is demonstrated and explained in the next section. in Table 2. The optimized circularly polarized antenna was then fabricated to experimentally validate the proposed design which is demonstrated and explained in the next section. in Table 2. The optimized circularly polarized antenna was then fabricated to experimentally validate the proposed design which is demonstrated and explained in the next section.

Measurement Results
The optimized 5G NR band circularly polarized antenna was finally fabricated to measure the experimental results. The photographs of the proposed prototype are presented in Figure 11. The closeup, front view, top view, and back view of the antenna are depicted in Figure 11a-d, respectively. The ECCOSTOCK HiK with permittivity, ε r = 10 having loss tangent (δ) of 0.002 was used as DR material. An 80 × 80 mm aluminum ground plane was used. The feeding strip was cut from adhesive copper tape to easily stick to the surface of the DRA. An SMA was soldered at the feed point at the center of the DRA with the feeding strip. S 11 was measured using a vector network analyzer (VNA), while far-field parameters were measured in the anechoic chamber. A 50-Ω coaxial cable was used to connect the SMA with the VNA. Double-sided copper tape was used to stick the DRA on the aluminum ground plane] to remove the possible air gap, according to the procedure explained in [26].
The resonant mode frequencies of the RDRA could be predicted using mathematical equations of the dielectric waveguide model, as explained in [27]. According to DWM TE x δ13 was estimated at 3.89 GHz and TE y 1δ1 at 4.53 GHz. The comparison between simulated and measured S 11 of the CP DRA is presented in Figure 12. TE x δ13 was simulated at 4.13 GHz and measured at 4.12 GHz while TE y 1δ3 was simulated at 4.52 GHz and measured at 4.51 GHz. A close comparison was observed between predicted, simulated, and measured values. The comparison of these values is presented ed in Table 3. The impedance matching bandwidth (|S 11 | ≤ 10 dB) expanded from 4.05-4.81 GHz in simulation and 4.01-4.83 GHz in measurement. The antenna provided a measured S 11 over a bandwidth of~17%. A small marginal difference between simulation and measurement was observed, due to cable losses and measurement errors.

TE
4.53 4.52 4.51 Figure 13 shows the simulated and measured CP response of the proposed RDRA in the boresight direction (i.e., Ɵ = 0 , Φ = 0 ). As presented, the axial ratio extended from 4.4-4.84 GHz in simulation, while 4.38-4.82 GHz was measured during the experiment. Circular polarization of bandwidth of ~10% was achieved, both in simulation and measurement. The minimum AR value was simulated at 4.44 GHz and measured at 4.43 GHz with a magnitude of 1.92 dB and 1.4 dB, respectively. The negligible difference between theory and experiment was due to cable losses and measurement imperfections. The successful overlap of S11 and AR bandwidths is depicted in Figure 14. As shown, the overlapped bandwidth was good enough to cover the targeted 5G NR Sub-6 GHz band i.e., n79 . A stable and satisfactory performance was offered by the antenna during the experiment.  . Simulated and measured S11 of the CP 5G RDRA. Figure 12. Simulated and measured S11 of the CP 5G RDRA.  Figure 13 shows the simulated and measured CP response of the proposed RDRA in the boresight direction (i.e., θ = 0 • , Φ = 0 • ). As presented, the axial ratio extended from 4.4-4.84 GHz in simulation, while 4.38-4.82 GHz was measured during the experiment. Circular polarization of bandwidth of~10% was achieved, both in simulation and measurement. The minimum AR value was simulated at 4.44 GHz and measured at 4.43 GHz with a magnitude of 1.92 dB and 1.4 dB, respectively. The negligible difference between theory and experiment was due to cable losses and measurement imperfections. The successful overlap of S11 and AR bandwidths is depicted in Figure 14. As shown, the overlapped bandwidth was good enough to cover the targeted 5G NR Sub-6 GHz band i.e., n79 (4.4-4.8 GHz). A stable and satisfactory performance was offered by the antenna during the experiment.
Sensors 2022, 22, x FOR PEER REVIEW 13 of Figure 14. Simulated and measured S11 and AR overlapping bandwidths of the CP 5G RDRA.
The simulated and measured radiation patterns of the proposed CP 5G RDRA a presented in Figure 15. The radiation patterns were computed and measured at thre different frequencies. The antenna provided stable radiation patterns with left-han circular polarization, since the left-hand field component was greater than the right-han field component by a margin of more than 20 dB at minimum AR frequency i.e., 4.43 GH as shown in Figure 15b. The right-hand CP response could be achieved by reversing th feeding strip. The simulated and measured gain of the CP antenna are shown in Figu  16. The antenna offers a useful gain of ~6.2 dBic throughout the CP bandwidth. This hig gain was achieved by excitation of the higher-order mode. A reasonable resemblance wa observed between simulated and measured results. The simulated and measured radiation patterns of the proposed CP 5G RDRA are presented in Figure 15. The radiation patterns were computed and measured at three different frequencies. The antenna provided stable radiation patterns with left-hand circular polarization, since the left-hand field component was greater than the right-hand field component by a margin of more than 20 dB at minimum AR frequency i.e., 4.43 GHz, as shown in Figure 15b. The right-hand CP response could be achieved by reversing the feeding strip. The simulated and measured gain of the CP antenna are shown in Figure 16. The antenna offers a useful gain of~6.2 dBic throughout the CP bandwidth. This high gain was achieved by excitation of the higher-order mode. A reasonable resemblance was observed between simulated and measured results.    The simulated and measured radiation patterns of the proposed CP 5G RDRA are presented in Figure 15. The radiation patterns were computed and measured at three different frequencies. The antenna provided stable radiation patterns with left-hand circular polarization, since the left-hand field component was greater than the right-hand field component by a margin of more than 20 dB at minimum AR frequency i.e., 4.43 GHz as shown in Figure 15b. The right-hand CP response could be achieved by reversing the feeding strip. The simulated and measured gain of the CP antenna are shown in Figure  16. The antenna offers a useful gain of ~6.2 dBic throughout the CP bandwidth. This high gain was achieved by excitation of the higher-order mode. A reasonable resemblance wa observed between simulated and measured results.  In Table 4, the proposed CP 5G RDRA is compared with recently reported RDRAs in the literature. It can be concluded that the proposed design configuration offers wide CP bandwidth achieved by excitation of the orthogonal higher-order mode pair, with very simple design geometry. Basic DR shape, i.e., rectangular, was used as radiating element along with the implementation of a simple feeding mechanism. Moreover, the field distribution inside the DRA was controlled by short adjacent magnetic dipoles. The spacing between these dipoles was responsible for the gain of the DRA, which could be improved by increasing the spacing. The spacing could be enhanced by excitation of the DRA in higher-order mode [28]. In literature, different efforts have been made to excite the RDRA in higher-order mode for high gain applications but the reported geometries are complicated and not easy to implement. In the proposed antenna higher-order orthogonal mode was excited using a new conformal feeding strip that was cut from an adhesive copper tape, which is a simple and cost-effective solution to the problem. In Table 4, the proposed CP 5G RDRA is compared with recently reported RDRAs in the literature. It can be concluded that the proposed design configuration offers wide CP bandwidth achieved by excitation of the orthogonal higher-order mode pair, with very simple design geometry. Basic DR shape, i.e., rectangular, was used as radiating element along with the implementation of a simple feeding mechanism. Moreover, the field distribution inside the DRA was controlled by short adjacent magnetic dipoles. The spacing between these dipoles was responsible for the gain of the DRA, which could be improved by increasing the spacing. The spacing could be enhanced by excitation of the DRA in higher-order mode [28]. In literature, different efforts have been made to excite the RDRA in higher-order mode for high gain applications but the reported geometries are complicated and not easy to implement. In the proposed antenna higher-order orthogonal mode was excited using a new conformal feeding strip that was cut from an adhesive copper tape, which is a simple and cost-effective solution to the problem

Conclusions
This paper reported on a new low-cost circularly polarized RDRA for 5G NR Sub-6 GHz band applications. A new conformal metal strip was utilized for excitation of the higher-order orthogonal mode pair to generate CP response. A circular polarization over a bandwidth of~10% was achieved along with a wide impedance matching bandwidth of 17%. The axial ratio bandwidth was in conjunction with impedance matching bandwidth. A left-hand CP response was achieved with stable radiation patterns throughout the circular polarization bandwidth. A useful gain of~6.2 dBic was attained by the antenna. The use of a simple and low-cost feeding mechanism, i.e., a unique conformal strip, to generate a circularly polarized wave by excitation of higher-order orthogonal modes, is the significant feature of this research, as compared to reports in the literature. A good agreement was observed in simulation and experimental results.