A Novel Circular Reconfigurable Metasurface-Based Compact UWB Hybrid Coupler for Ku-Band Applications

A novel circular reconfigurable metasurface (MS) based compact ultra-wideband (UWB) hybrid coupler is developed for Ku-band applications. The coupler is developed using the substrate-integrated gap waveguide (SIGW) technology. The coupler structure consists of two layers, the bottom layer represents the artificial magnetic surface of the periodic structures and the ridges in between that guide the wave in the required direction with minimum dispersion. It involves the coupling section with a centered etched slot and two additional vias to achieve the basic hybrid coupler properties. This layer is nominated as the ridge layer. The second layer is a circular shape of a dielectric gap loaded with the top ground. The top ground is left solid for a non-reconfigurable coupler. Concerning the reconfigurable coupler, this layer contains an artificial metasurface of Jerusalem cross elements where the copper is etched around. This layer is nominated as the gap layer. This MS surface is mechanically rotated to offset the magnitude and phase of the signal going to the through and coupled ports. The findings obtained from the simulations show that the reconfiguration can be accomplished by rotating the MS around the source coupler’s central axis. The rotation is tested between 0° to 180° in the counter-clockwise direction. The operating frequency range of the coupler is between 11.94 to 16.91 GHz, which covers approximately the whole Ku-band. The coupler delivers continuously adjustable amplitude between 2.6 and 4.8 dB while the phase differences within 77° to 105° over a fractional bandwidth (FBW) of 34.45%. It is manufactured using PCB technology and measured using network analyzer. A strong agreement is achieved between simulations and measurements. The proposed coupler can be used in traditional beam-forming and beam-steering networks by changing the rotation angle or the operating frequency. The developed coupler can replace the Butler and Bless matrices with their complication, heavy number of phase shifters, and crossover problems. The current work can be extended to operate in the mm-Wave band by changing the dimension and the material of the unit cell of the ridge layer of the coupler.

acquisition, are vital for 5G and future 6G communication systems, making directional couplers critical components. At low-frequency bands, waveguides and microstrip couplers have achieved remarkable advancements in recent years. However, microstrip couplers encounter significant dielectric losses in the millimeter wave frequency region. In addition, standard metallic waveguide couplers suffer from difficulty in production, high costs and the ability to integrate with other networks. The operation at higher frequencies requires a suitable waveguide. Thus, gap waveguide (GW) technology has emerged as a potential replacement that could be employed in such bands without any of the aforementioned restrictions [1], [2]. As a result, many coupler designs have been studied using GW technology in recent years [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16]. However, none of them is reconfigurable.
The demands for reconfigurable devices are increasing because of the great flexibility and simple expansions they offer for new usage conditions. Moreover, the ability to reconfigure a wireless communication system has become a necessity to save costs and provide the best performance for various applications. Couplers are commonly used to build a beamforming network and produce the required radiation beams. A hybrid coupler, acting as a phase shifter, is also capable of performing a power division function and a phase shift. It is important to note that a typical coupler-based beamforming network can only produce limited beams because of the standard phase differences such as 90 • and 0 • /180 • . But, it is typically required to have wide constantly controllable phase differences for greater flexibility. The easiest approach is to add a phase shifter after the power divider or the hybrid coupler. However, the following three major issues arise later: high insertion loss, significant phase change, and enormous circuit size [17]. The radiation beams can be constantly scanned over a large area if the coupler can give a reconfigurable phase difference. That is why it is essential to have a coupler with a broad phase adjustment range.
Reconfigurable couplers have garnered a lot of interest in microwave integrated circuits with variable power-dividing ratios [18], [19], [20], [21], and operating-frequency adjustment [22], [23], [24], [25]. Only a few numbers of structures have been documented in the literature since the reconfigurability in phase difference is more difficult than that in frequency or power division ratio [26], [27], [28], [29]. Varactor-tuned couplers were investigated to complete the continuously adjustable differential phase at a given frequency in order to increase the phase coverage. Generally, tuning ranges between 45 • and 135 • can be provided with the current phase-reconfigurable couplers. The periodic employment of varactors as transmission line loads allows control voltage modification of the equivalent circuit capacitance which supplies a 30 • -150 • phase difference [26]. In [27], a varactor-loaded branch line coupler ensures equal power division with a configurable phase difference of 45 • to 135 • . The implementation of an adjustable phase shifting unit for the horizontal branch can cause the phase difference to be continuously set from 45 • to 135 • [28]. In [29], a tunable unit was made up of open and shorted stubs, tunable capacitors, and two eighth wavelength linked lines which achieved a 0 • to 180 • phase difference. All the reported phase tunable capacitor designs are lumped elements based and their operating frequencies are mostly below 3 GHz.
In this study, a novel mechanically reconfigurable compact SIGW-based hybrid coupler is presented. A variation of adjustable amplitudes and phases are achievable by rotating the MS surface of the top layer.
The paper is arranged as follows: In Section II, the non-reconfigurable and MS tunable couplers are presented. In Section III, the performances of both couplers are discussed based on the experimentally validated results. Section IV concludes this article.

A. SIGW DESIGN
Basically a gap waveguide is developed between two parallel conducting metallic plates, one of which has a textured surface made of pins or vias to produce a perfect magnetic conductor (PMC) surface. The other layer has a perfect electric conductor (PEC) on top of the structure. These two layers are separated by a gap height that is less than a quarter wavelength of the PMC plate. As a result, a stop band that prevents the parallel-plate modes from propagation can be achieved in the PMC-PEC cut-off zone. Metal ridges are interspersed between the textured plates to sustain the wave in a certain direction. The waveguide is created in a small space between the ridge and the top metallic plate, which is typically filled with air but in our work it is filled with a dielectric material which is known as substrate-integrated waveguides (SIGW). The SIGW has more advantages over ridge gap waveguide (RGW) with an air gap being replaced by a dielectric material as it maintains a constant gap height that minimizes the risk of possibility that this gap collapses in the presence of stress or impact. The fabrication process is also simpler (PCB-based). Another disadvantage of the RGW with an air-filled structure is that it requires very accurate CNC machining and its performance is unstable. Inside the gap, a quasi-TEM condition is maintained over the ridge and an electromagnetic (EM) leakage is stopped [30], [31]. So, our design is carried out based on SIGW technology.
Based on SIGW technology, the unit cell of the current design is composed of two parallel layers positioned as a bottom Rogers RO4350B ( r = 3.66) and a top Rogers RO4003C ( r = 3.55) substrates. In the bottom layer, the ground and the conductor mushroom patches are connected by vias. The directing ridge is positioned in the middle of the unit cells to activate the desired propagating quasi-TEM mode.   that the ridge lowers the stopband by approximately 5.6%. Fig. 2 demonstrates the feeding network of the proposed SIGW and the surface current distribution for two distinct frequencies, one is in the band and the other is out of the band. At 15 GHz, it is evident that the wave is constrained while leaking occurs at 5 GHz which is out of the band. Table 1 shows all the dimensions of the unit cell and supercell structures. It is determined that 1.6 mm is the optimal ridge width value. In essence, by adapting the strip line impedance equation, the characteristic impedance of the ridge can be determined. A small deviation from the results is predicted because the structure consists of two different substrates with close r values.
For the measurement of the scattering parameters of the coupler, a standard RF edge SMA connector should be connected to the ridge which is difficult to be implemented. A transition shown in Fig.3 is designed for this purpose. A microstrip line of characteristic impedance of 50 connects the ridge line to the SMA connector. The microstrip line is implemented on the top layer with a thickness of 0.45 mm and it contacts the upper surface of the ridge as shown in Fig. 3. Table 2 shows the microstrip transition dimensions.    The bottom layer of the coupler is illustrated in Fig. 4, it contains four reciprocally coupled printed ridge gap waveguide (PRGW) lines. The coupling section is a circular junction patch with a 45 • elliptical slot in the center of those lines. In addition, two more vias are added at an orthogonal angle to the slot axis. This slot and vias are intended for achieving a better power distribution in 3 dB hybrid coupler. This coupler's design is comparable to a typical microstrip or bulky wave couplers. According to the basic design concept [5] of the hybrid coupler, if two parallel lines have characteristic impedances of Z 0 , the other two parallel lines should The lengths L1 and L2 are not VOLUME 10, 2022  precisely identical to the λ g /4 value, two correction factors are provided for designing the SIGW coupler effectively as  given by Eqs. 1,2.
where b and c represent two additional correction factors.
The starting values of b and c can be set to zero, but the final values must be tuned depending on the optimal passband and the isolation performance. The optimum values for L1 and L2 are 3.1 mm and 5.8 mm, respectively. The elliptical slot's location specifies the isolation port to be used. The isolation is port 2 if the slot is slanted to the left; if it is inclined to the right, the isolation is port 4. Fig.5 demonstrates the electric field distribution in terms of elliptical slot placement. In order to improve the power distribution, two vias were also placed at the bottom of the circular patch. The optimized dimensions of the elliptical slot are d1 = 1.8 mm and d2 = 4.6 mm and the diameter of the circular patch    Fig. 6. It can be seen that the center frequency shifts depending on the size of the slot. The optimized values of d1 and d2 are obtained by changing one parameter while keeping the second one constant and vice versa. If d1 and d2 decrease, the center frequency shifts to higher frequencies. If d1 values increase, the shifting occurs at lower frequencies. An increase in the d2 does not affect the center frequency. The effects of additional vias in term of power division are shown in Fig.7. It is evident that huge losses occur between 14 GHz   and 15 GHz without vias. Adding two extra vias helps in the equal power division. It is important to point out that this structure can be easily extended to operate in the mm-Wave band for 5G/6G applications. This can be achieved by the precise design of the unit cell of the ridge layer of the coupler as depicted in Fig. 8. This dispersion diagram shows that the artificial magnetic conductor (AMC) vias can generate a bandstop to accommodate a quasi-TEM mode over two subbands of mm-Wave applications with a BW more than 18 GHz. The material layer is changed to Rogers RT5880 and the dimension is changed according to the values in Table 3. It is considered for the future work of the current structure.     The basic formulas for the MS's equivalent impedance Z, the refractive index n, µ r , and r can be written as [32]: (3) where k 0 , d are the wave number and MS equivalent thickness, respectively. r = Re{n/Z } and µ r = Re{nZ }.
The scattering parameters and the normalized absorption rates are shown in Fig.10. One can see that the MS behaves as a reflector over the whole operating frequency range except between 16 and 17 GHz, it is an absorber. In the frequency range from 16.4 to 16.7 GHz, the MS is a left-hand metasurface with a negative index of refraction n, negative µ r , and negative r values. It is a right-hand gap absorber from 16.7 up to 17 as shown in Fig. 11.
The proposed reconfigurable coupler design is based on Jerusalem cross MS as depicted in Fig.12. The top layer of the SIGW is circularly cut for the MS to be mechanically adjusted and rotated. The diameter of the rotating surface is 20 mm.

C. FABRICATED COUPLER
This section is devoted to the fabrication of the proposed coupler using PCB technology. Concerning the coupler structure, it is divided into two main layers. The gap layer includes the dielectric gap and top ground. The ridge layer comprises the vias, the ridge and the feeding network including the transition as depicted in Fig. 13. For the sake of reconfiguration the top ground of the gap layer includes a separate interior circular section of the MS structure which can be rotated for different angles as shown in Fig.14. Fig. 15 illustrates the measurement setup for measuring the scattering parameters of both couplers using network analyzer ROHDE & SCHWARZ ZVB20.

III. RESULTS AND DISCUSSION
Concerning the non-reconfigurable coupler, Fig. 16 depicts the simulated and measured scattering parameters of the proposed design. One notices that it operates 12.14 to 15.4 GHz under -10 dB level. Also, the coupling coefficients at port 3 and port 4 are nearly 3 dB with isolation at port 2. Fig. 17 shows the phases of the scattering parameters at ports 3 and 4. One can infer that the phase difference between them is nearly 90 • . It is clear that the proposed coupler is a SIGW-based technology that achieves the hybrid parameters performance of the conventional couplers. The strong agreement between the measured results and the simulated ones validates all the new findings of the study.
Concerning the proposed reconfigurable coupler, the top MS layer is mechanically rotating in counter-clockwise directions to achieve different scattering parameters at ports 3 and 4. As shown in Fig.18 the reflection coefficient values vary with rotation keeping a persistent common frequency band. The widest bandwidth (UBW) occurs between 11.94 GHz to 16.91 GHz at 150 • while the narrowest one is between 13.07 GHz to 16.34 GHz at 30 • . The common BW among the different angles of incidence is approximately 4.2 GHz. The return loss is almost below -20 dB for all rotation angles. The isolation is greater than 15 dB for all cases. As can be observed, the two-output amplitude responses are slightly changing according to the rotation angle. The output power remains steady between 2.6 dB and 4.8 dB range in the middle as shown in Fig 19. However, the power fluctuates in port 4 at 8 dB at the corner of the operating frequency. As demonstrated in Fig. 20, one can notice that the phase variation at ports 3 and 4 cover the range from 180 • to -180 • keeping the phase difference between them ( S 41 -S 31 ) tunable in the range from 77 • to 105 • . The maximum phase  difference of 105 • is attained with the MS at 0 • position, while a rotation of 90 • achieves the lowest phase difference of 77 • . Various rotation angles can be employed to produce different phases in the range from 77 • to 105 • . Table 3 provides an extensive comparison of the proposed design with previously reported tunable couplers. Firstly, to the best of the authors' knowledge, our proposed design is the first adjustable phase coupler in the Ku band/mm band ranges. That is why we compare our work with reconfigurable couplers in different frequency ranges. Our work supplies the widest FBW compared to other studies. For example, GW-based hybrid couplers reported in [13], [14], and [15] provide a fixed phase difference. The currently proposed coupler design can provide a narrower phase tuning range in comparison to the previously reported ones [26], [28], [29]. These authors used many varactor diodes to be able to tune the phase. However, using this type of lumped element causes huge losses at high frequencies. The frequency reconfigurable coupler in [22] employed an MS to adjust the frequency however the phase difference was kept fixed. Finally, our obtained results show that the additional insertion loss of the proposed coupler is minimum compared to the other studies. It is maintained within the acceptable range.

IV. CONCLUSION
In this paper, a novel reconfigurable MS-based small UWB hybrid circular coupler is developed utilizing the SIGW technology. To the best of the authors' knowledge, our proposed design is the first adjustable phase coupler in the Ku band/mm band ranges. The findings obtained from the simulation show that phase reconfiguration can be accomplished by rotating the MS around the source coupler's central axis. The proposed design is continuously providing a variable phase difference over a fractional bandwidth of 34.45% between 77 • and 105 • . It is manufactured using PCB technology and measured using a network analyzer. The findings from simulation and measurement exhibit good agreement. This novel proposed coupler can be best used as a traditional beamforming network. It can also be used for beam steering by changing the rotating angle or the operating frequency. So, it could be a good candidate for radar applications. Moreover, it could also be used for generating the well Grid Of Beams (GOB) networks. The current design supplies the widest FBW compared to other studies. It is very important to point out that our design provides a simple, easy fabrication and implementation, compact in size, and easy-to-integrate solution with array antennas etched on the top layer of the structure as slot antennas without adding extra PCB materials to construct the antenna array. These novel features are solely comparing our design with all others mentioned in the literature. Also, the amplitude variation can be adjusted to change the beamwidth of the generated beams in the beamforming network. This work is being extended by changing the unit cell material and dimensions in the mm-Wave band. He published more than 100 conference and journal papers. His research interests include RF and microwave technology, antenna design, smart antennas for vehicles and radar sensors, satellite communication, and diagnoses of censer based on the electromagnetic properties of human organisms. He did a lot of applied research in RADAR absorbing materials including Chiral, FSS, and textile materials.
DIAA E. FAWZY (Member, IEEE) was born in Egypt, in 1968. He received the Ph.D. degree from Heidelberg University, Germany, in 2001. He was involved for many years in the German GSM-R Project and was responsible for the network planning and optimization. He has been with the Izmir University of Economics, since 2008, where he is currently a Professor and the Head of the Department of Aerospace Engineering. He is a reviewer in many international journals. His current research interests include microwave devices, computational electromagnetics, antennas design, mm-waves, MW technology, remote sensing, and AI.