Ultra-Wideband Hybrid Magneto-Electric Dielectric-Resonator Dipole Antenna Fed by a Printed RGW for Millimeter-Wave Applications

Future communication standards have an increasing interest in Millimeter Wave (mm-wave) bands, where wide bandwidth and secured communication can be assured. This trend in communication standards stimulates the research community to provide novel antenna configurations in the mm-wave bands. This article proposes a novel design and analysis of hybrid magneto-electric dielectric-resonator dipole antenna that features an electrically small size with ultra-wideband operation and consistent radiation characteristics in the mm-wave band. The proposed antenna is designed based on the combination of multi-resonances produced by a Magneto-Electric (ME) dipole and stacked Dielectric Resonator Antennas (DRAs). In addition, the proposed antenna is implemented using state-of-the-art Printed Circuit Board (PCB) technology, namely, Printed Ridge Gap Waveguide (PRGW) for low loss and cost-efficiency. The combination between the ME-dipole and stacked DRA is adopted to ensure symmetric radiation characteristics in both E- and H-planes over ultra-wideband operation with a deep matching level. Both DRA and ME-dipole elements are designed and studied separately, where a systematic design procedure is presented to obtain initial design parameters. Proper integration between the radiating elements introduced an electrically small size antenna ( $0.64\,\,\lambda \times 0.48\lambda $ ) covers the full Ka-band (26-40 GHz) with a matching level beyond −20 dB and gain stability of 8± 1 dB. The antenna prototype is fabricated, where a good agreement is shown between both simulated and measured results.

Low-profile printed antennas are essential building blocks for modern communication systems in mm-wave bands, where both cost and size are critical features. The printed antennas have been addressed extensively based on either classical guiding structure such as microstrip lines [9]- [12] or modern technology such as substrate integrated waveguide (SIW) [13]- [16]. Although the classical guiding structurebased antennas can provide the required bandwidth and compactness with a cost-effective solution in low-frequency bands; however, the dielectric and radiation losses are significant and have more impact at higher frequency bands, which limits their applications at mm-wave bands. On the other hand, the radiation losses can be significantly reduced through the utilization of SIW technology, where the signal is fully confined between two metallic plates and guided from both sides by periodic vias. However, dielectric losses represent a huge drawback in this technology as the signal propagates inside the dielectric region. Another promising structure is the printed ridge gap waveguide (PRGW), where the radiation and dielectric losses are reduced through the propagation in an air gap above the ridge surrounded by periodic electromagnetic band-gap cells [17]- [19]. This state-of-the-art guiding structure has attracted great interest from the research community especially in mm-wave bands introducing various microwave components and antenna configurations. However, there is still a significant room of improvement in the performance of these components to overcome the associated challenges reported in the literature.
Recently, various mm-wave PRGW based antenna designs have been reported in the literature, which is separately deploying either ME-dipole or DRA configurations. Each of these antenna types has its advantages which include a wide bandwidth and symmetrical E-and H-plane stable radiation characteristics for ME-dipole [8], [23] or electrically small size and high efficiency for DRA configurations [11], [12]. For example, some presented antenna structures have sacrificed the bandwidth to provide a high gain solution [8], [20]- [22]. Others proposed antenna configurations that cover a wide bandwidth, nonetheless, with unstable gain [23], [24]. Moreover, most of the published configurations that have deployed dielectric resonator structures do not maintain symmetrical radiation characteristics over a wide operating bandwidth [25]- [27]. Therefore, introducing a novel type of mm-wave antenna incorporating both ME-dipole and DRA features is essential to elevate the performance of future wireless communication systems.
In this article, we propose the design and analysis of a novel Hybrid Magneto-Electric Dielectric-Resonator Dipole (ME-DRD) antenna based on PRGW technology for mmwave applications. The proposed antenna achieves an outstanding performance due to the combination of the two fundamental types of radiators, namely, ME-dipole and DRA. As a result, the proposed antenna is achieved not only an ultrawide bandwidth, identical E/H-plane, stable gain, compact size but also a deep matching level (<-20 dB) compared with other published designs. The main radiators for the proposed antenna are well studied and designed separately, where the dimensions of the geometrical configuration parameters are calculated based on the required resonance frequency through the operating bandwidth. Proper integration between the proposed antenna elements was performed, where the proposed ME-DRD antenna is tuned to achieve a deep matching level below −20 dB covering the frequency range of 26-40 GHz. In addition, the proposed antenna configuration produces symmetric radiation characteristics over the suggested bandwidth. Furthermore, a stable gain of 8±1 dB is achieved over the operating bandwidth with a compact size of (0.64 λ ×0.48λ ). To the best of the authors' knowledge, the proposed antenna configuration outperforms the published antenna structures in the literature, where a detailed comparison between this work and other related published contributions is addressed in detail.
This article is organized as follows: In Section III, the proposed structure is presented along with the analysis of the antenna modes. This is followed by a detailed discussion on the fabricated prototype, where the comparison between the measured and the simulated response is criticised in Section IV. In Section V, the performance of the antenna is evaluated through a comparison with other related published work to highlight the proposed antenna edge. Finally, the outcomes of this work is summarized in Section VI.

II. PROPOSED PRGW STRUCTURE DESIGN AND EXCITATION
The proposed antenna consists of PRGW antenna connected to a PRGW feeding line, where in a later section this feeding line will be connected to a standard connector for measurement purposes. The PRGW feeding line is designed with a cell size of P = 1.7 mm and a patch radius of r = 0.7 mm. The designed cell is printed on Roger RT 6002 substrate with a height of H F = 0.762 mm, while the via diameter is 0.38 mm. The proposed unit cell, as well as the straight PRGW feeding line, are shown in Fig. 1(a) and 1(b), respectively. These parameters are deployed to model the PRGW, where the simulated modes with the presence of the ridge are shown in Fig. 1(c). It can be depicted form this figure that a quasi-TEM mode exists in the frequency band from 23 GHz to 40 GHz, which is a bit larger than the suggested frequency band. Accordingly, this PRGW cell configuration provides a suitable guiding structure for the antenna design. The second part of the proposed transition is a Microstrip line to PRGW transition, where the middle strip is connected to the printed ridge through a matching section as shown in Fig. 2(a). The utilized matching section is simple and effective, where the microstrip line is connected to a dielectric filled PRGW, which in turns, is connected to the air filled PRGW. The width of the intermediate section is used as a tuning parameter to achieve the required matching level. The simulated response of the entire transition is shown in Fig. 2(b), where a matching level better than −19 dB is achieved over the entire frequency band. In the next section, the proposed antenna design, fed by this PRGW, is discussed.

III. PROPOSED ANTENNA DESIGN
The proposed ME-DRDA antenna structure, shown in Fig. 3 consists of two basic elements: a ME-dipole and two-layered DRA loading. Both elements are selected to produce symmetrical radiation patterns, which is a mandatory condition to have a stable radiation characteristics over an ultra-wide bandwidth. It is worth mentioning that, it is hard to achieve a sufficient matching level if the proposed configuration is only fed by a slot deployed in the ME-dipole ground plane. Therefore, the integration between these radiators is fed through a PRGW line terminated with a wide bandwidth hexagonal-shaped open stub located in the bottom layer, as shown in Fig. 3. This section is divided into three subsections:the first two subsections discuss the two antenna basic elements, while the third subsection presents the integration and the realization methodology.

A. PRGW ME-DIPOLE ANTENNA DESIGN
The ME-dipole antenna is selected to be one of the main radiators of the proposed ME-DRD Antenna as it offers a superior radiation characteristics compared to other radiators, such as slot or electric dipoles antennas. The ME-dipole is an integration between a printed dipole and a slot simultaneously excited, where each component is responsible of a certain specific resonance. These two resonances can be adjusted to overlap each other and provide a wide bandwidth; however, it is not sufficient to cover, for example, the whole Ka-band with a reasonable matching level.
Although the printed ME-dipole antenna design is presented several times in the literature, the design procedures have not been described in details. To get an initial estimate for the proposed ME dipole dimensions shown in Fig. 4(a), we assume a dipole arm length of quarter wavelength and a slot length of half wavelength. As a result, the dimensions of both the dipole arm and the slot can be written as follow: where, C is the velocity of light, f ol and f oh are the lower and higher resonance frequencies, respectively, ε r1 is the dielectric constant of Roger RT6002, and ε r1eff = ε r1 +1 2 is the ME-dipole effective dielectric constant. Afterwards, the dimensions are selected to allocate the two resonances at 28 GHz and 40 GHz, respectively, where the proposed ME-dipole return loss is shown in Fig. 4(b). For further clarification on the operation of the proposed ME-dipole, both electric and current distributions are plotted in the same figure. It can be noticed that current distribution is concatenated on the ME-dipole patches which represents the electric dipole, while the electric field is distributed along the 2030 VOLUME 10, 2022 gap between the patches, which confirms the magnetic dipole operation.

B. PRGW DRA ANTENNA DESIGN
The other radiator adopted to improve the performance of the proposed ME-DRD antenna is the DRA as it has smaller size and provides more degrees of freedom as many as available low-cost material products. In addition, it is suitable for mm-wave frequency bands as it is featured with a high efficiency due to the absence of metallic surfaces. The loaded antenna with the DRA is illustrated in Fig. 5(a), which basically consists of two cascaded dielectric resonators on the top of the slot antenna. This DRA configuration is an aperture coupled through the same ME-dipole slot to excite the fundamental mode TE 111 . The DRA loading is responsible of the third and fourth resonances to further widen the operating bandwidth. These resonances are chosen to occur in between the ME-dipole resonances to ensure a sufficient matching level over the operating bandwidth. The resonant frequency of the fundamental mode f odr1,TE 111 can be calculated for the lower DRA as follows: where, C is the speed of light and k o is the free space wave number. An effective dielectric constant for the proposed DRA antenna is calculated as the DRA is placed on the ME-dipole substrate which has different dielectric constant. Therefore, the effective dielectric constant ε r2eff is calculated VOLUME 10, 2022  as follow: where, H t = H DRl + H ME , ε r1 and ε r2 are the dielectric constant of Rogers RT6002 and RT6035, respectively. Equations (3-7) are used to estimate initial values of the dimensions of the dielectric resonators in order to resonate around 26 GHz and 34 GHz for the upper and lower DRAs, respectively. The response of the proposed DRA is shown in Fig. 5(b), which highlights a potential resonance at 34 GHz for a single DRA, and dual resonances at 26 GHz and 34 GHz for a stacked DRA as shown in Fig. 5(c). However, as can be observed in Fig. 5(b), by adding the upper DRA, the resonance frequency of the lower DRA is slightly shifted to 36 GHz. In addition, the side view of the simulated E-field distribution of the propose DRA is depicted in Fig. 5(b), where the field distribution is similar to the ideal TE 111 mode of the rectangular DRA.

C. INTEGRATION AND FABRICATION CONSIDERATIONS
The integration between both main radiators is carried out through staking both structures with the initial dimensions. Based on several parametric studies, we found that the configuration of larger DRA on the top of smaller one results in a good matching level for the proposed antenna rather than the opposite direction. Gradually increasing the aperture size will allow a smooth transition or radiation of the electromagnetic wave to the free space. In addition, this will increase the gain of the antenna as we will have a larger aperture. The initial response is promising where a matching level below −10 dB is achieved over the operating bandwidth as illustrated in Fig. 6. Hence, an optimization process is performed to obtain a sufficient matching level with a stable gain. The response after optimization for both the gain and the return loss is shown in Fig. 7, where a matching level below −20 dB with a gain stability of ±1 dB is achieved over the whole Ka band. It is worth to mention that the adopted technique to design the proposed antenna results in a deep matching level (<−20 dB) over the whole Ka-band, which is higher than any ME-dipole or DRA that reported in the literature, where −10 dB is the achieved matching  level. Thanks for the combination between the ME-dipole and DRA, which results in a novel antenna with superior characteristics. The realization of the proposed antenna is performed through inserting the high dielectric constant material forming the DRA inside a perforated substrate as shown in Fig. 3. This is deployed to ensure the alignment through the fabrication process. The relative permittivity of VOLUME 10, 2022 the holding substrate is calculated as follow: where, ε r4 is the dielectric constant of Rogers RT 5880 LZ. Table 1 indicates the final optimized dimensions of the proposed antenna after the optimization process.

IV. FABRICATION AND EXPERIMENTAL VALIDATION
The proposed ME-DRD antenna structure shown in Fig. 8() is fabricated using the conventional PCB process, and combined together using epoxy at high temperature and pressure [28]. A standard SouthWest Microwave 2.92mm connector is used to test the fabricated prototype as shown in Fig. 8(b). The prototype is measured for the return loss and gain, where a good agreement is achieved between the simulated and the measured response in Fig. 8(c). The proposed antenna achieved a measured matching level of −13 dB over a 42% relative bandwidth at 33 GHz that covers the whole Ka-band. However, a slight discrepancies can be observed for the measured results, which is probably due to fabrication tolerances. More specifically, the epoxy used to assemble the layers, and the soldering that used to ensure the electric contact between the ridge line and the microstrip line. Another factor that may cause the difference between the simulation and the measurement is the changing of the materials dielectric constants at mm-wave frequency compared with the written in the substrate data sheet that are evaluated at 10 GHz. The radiation characteristics of the fabricated antenna are measured inside an anechoic chamber, where the measured setup is shown in Fig. 9. The gain of the fabricated antenna is measured, where a good agreement compared to the simulated results can be observed in Fig. 9(b). The proposed antenna provides a gain of 8 dBi ± 1 dBi in the whole operating frequency band. Moreover, the radiation patterns were measured and compared with the simulated ones at three different frequencies as shown in Fig. 10. This figure demonstrates that both simulated and measured results are in a good agreement, where the discrepancy between the results is caused by the inaccuracy or misalignment in the radiation pattern measurement setup. The proposed antenna has a symmetrical radiation pattern with a sidelobe level less than −15 dB and a low cross-polarization level less than −20 dB are achieved over the whole frequency band. The radiation pattern is symmetrical in both E-and H planes and stable, which are the main characteristics of such novel hybrid technique that used in this work.

V. PERFORMANCE EVALUATION
Multiple designs have been proposed based on ME-dipoles or dielectric resonators at mm-wave bands in the literature. Several articles have introduced ME-dipole to cover wideband operation around 30 GHz based on printed ridge gap waveguide, packaged microstrip line, and substrate integrated gap waveguide technologies [19], [21], [23]. The bandwidth of such designs is around 20% with a gain in the range of 5 to 8 dB with large variation over the operating bandwidth. Other designs enhanced both the gain and the bandwidth through loading the ME-dipole with Split Ring Resonators (SRR) or PMC surfaces [8], [20], [22], [24]. These techniques can widen the bandwidth up to 40 % and increase the gain by 2 dB. However, the size of the antenna and gain variation increase significantly, which represent obvious drawbacks to these techniques. In addition, most of these designs are using inverted microstrip line that requires an insertion of a substrate layer above the mushroom surface, which adds more space between the mushrooms and the upper ground. Therefore, the dielectric loss is scientifically increased compared with the case of printed RGW used in the proposed work. On the other hands, various antenna designs based on dielectric resonators have been proposed in mmwave bands, where most of these designs have a narrow bandwidth [25]- [27]. The proposed design in this work gives an excellent compromise, where the bandwidth is featured among all published work in a very compact size. In addition, the presented design provides a stable gain over the operating bandwidth. Furthermore, the proposed ME-DRD antenna is achieved a deep matching level compared with other published designs. This comparison is summarized in Table 2, where the gain, size, bandwidth and the host guiding structure are listed.

VI. CONCLUSION
In this paper, the design and analysis of a novel antenna configuration, namely, ME-DRD antenna fed by printed ridge gap waveguide in mm-wave frequency bands has been proposed. The proposed antenna has been designed through loading a slot with ME-dipole and stacked dielectric resonators to achieve an electrically small size and stable gain with ultra wideband operation. The printed ridge gap waveguide has been adopted to feed the antenna, where low loss and cost efficiency can be achieved. A systematic design procedure has been presented to obtain initial dimensions for the resonators used to construct the proposed antenna. The antenna prototype has been fabricated and tested, where a measured deep matching level and stable gain of 8 dBi ± 1 dBi over the whole operating bandwidth have been achieved. The obtained experimental results have shown a good agreement with those from simulation.With these promising features, the proposed antenna can be a good candidate for future wireless communication systems operating at millimeter-wave bands. The idea of this new type of antenna design could be extended to either mm-wave MIMO antenna systems or high-gain antenna arrays.