Abstract
A single-layer dual-band reflectarray cell is proposed in this work for future 5G systems. A reflectarray unit cell operating at 28/38 GHz is designed by adopting two pairs of miniaturized fractal patches, offering low losses (<0.7 dB) and almost full-phase ranges (≅320°) at both operating frequencies. The proposed configuration allows to achieve very small interelement spacings and negligible mutual coupling effects between the two bands, thus assuring an independent phase-tuning mechanism at both desired frequency bands. The designed compact cell is successfully adopted to demonstrate reflectarrays’ abilities in achieving fixed scanned-beam and/or multibeam patterns, under the dual-band operation mode. Full-wave numerical validations, performed on the synthesized reflectarray structures, confirm the effectiveness of the designed dual-band configuration in achieving independent radiation patterns and quite good bandwidths, at the two designed frequencies. Thanks to its compactness and versatility in achieving both frequency diversity and multibeam/scanned-beam radiation patterns, the proposed unit cell is appealing for future 5G applications.
1. Introduction
Nowadays, the development of new technologies for future fifth generation (5G) wireless communication networks is the main challenge in the telecommunications industry. 5G communication systems are expected to meet the growing demand for higher data rates (i.e., 1-10 theoretically gigabit per second (Gbps) [1]), as required by multimedia applications and the Internet of Things (IoT). To address this demand, 5G systems will use millimeter wave (mmw) frequencies, which represent one of the key enabling technologies in the development and implementation of 5G communication networks [1, 2]. However, the mmw frequencies are characterized by propagation limitations, such as higher path loss and shorter communication distances, mainly due to the atmospheric absorption of electromagnetic waves at higher frequencies [3]. For this reason, there is a need for designing high-gain antennas able to compensate for path losses. To this end, microstrip array antennas may represent a good candidate, providing also a narrow beamwidth and very thin profiles that are essential for 5G operations. Some interesting 5G array antennas have been recently proposed in [2, 4, 5].
A very attractive alternative solution for designing 5G antennas is offered by microstrip reflectarrays [6, 7]. They consist of an array of microstrip elements illuminated by a feed antenna. Each element is designed to compensate for the phase delay in the path coming from the feed and to introduce a phase contribution giving a prescribed main beam direction in the antenna radiation pattern. Thanks to their higher efficiencies, due to the adopted spatial feeding approach [6], reflectarrays represent a promising solution for designing high-gain/directivity 5G antennas. Furthermore, reflectarrays can be properly designed to offer several reconfiguration capabilities, which are very appealing for 5G systems [2], such as beam-steering functions and/or frequency agility [6, 8–11], multibeam radiation patterns [6, 9], and multiband operation modes [6].
In order to demonstrate reflectarray versatility in satisfying most of the 5G requirements, a single-layer dual-band reflectarray cell is investigated in this paper to operate within the Ka band (at 28/38 GHz), which is currently under consideration for 5G technologies [2]. The proposed cell is adopted to demonstrate reflectarrays’ abilities in achieving fixed scanned-beam and/or multibeam patterns, under the dual-band operation mode.
The fractal concept has yet to be applied in the literature to reflectarray antennas, as reported in [12–16]. In this paper, a modified layout of the fractal element, originally proposed by the authors in [12, 13], is investigated to design a compact single-layer dual-band cell with single linear polarization, offering low losses (<0.7 dB) and almost full-phase ranges (≅320°), at both operating frequencies. The miniaturization capabilities of the adopted fractal geometry are exploited to achieve a dual-band behavior, simply by embedding two pairs of miniaturized resonators within the same unit cell. Negligible mutual effects between the two bands are demonstrated, thus achieving an independent phase-tuning mechanism for each frequency band, by properly changing the fractal shapes. Unlike other multiband reflectarray cells presented in the literature [17–21], the proposed dual-band fractal cell allows to achieve the following: a simpler and thinner structure with respect to the multilayer stacked configurations [17, 18]; smaller unit cell size at both operating frequencies (≅0.4λ at 28 GHz, ≅0.54λ at 38 GHz) with respect to other single-layer configurations [19, 20], thus preserving the capability to point the main beam at large scan angles, without occurring in grating lobe phenomena; and smaller electrical interferences between the elements operating at the different bands [21], despite the adopted small interelement spacing.
As a first proof of the concept, the proposed cell is adopted to design reflectarrays able to independently scan the main beam at the two operating frequencies. A 1 dB gain-bandwidth of about 1 GHz is simulated at both operating frequencies, potentially offering several Gbps of data rates at 28 or 38 GHz [1]. Moreover, a dual-band multibeam reflectarray design is numerically demonstrated. Finally, the dual-band behavior of the proposed unit cell configuration is experimentally validated.
2. Dual-Band Reflectarray Unit Cell
The proposed dual-band unit cell is composed of two alternately arranged pairs of fractal patches (Figure 1), each designed to operate around a specific resonant frequency. The layout of the single patch is essentially derived from the 1st iteration fixed-length Minkowski patch originally proposed by the authors in [12]. The patch geometry, reported in Figure 1(a), is characterized by a beginning of dimensions (, 2). A smaller square of the side is removed from the center of the two lateral sides (i.e., the resonant sides), thus obtaining a linearly polarized element, along the -axis (Figure 1(b)). The scaling factor varies from 0 up to 0.45. The reflection phase tuning is realized by independently varying the fractal scaling factor of each element, leaving unchanged the patch size . The above fractal shape allows to fit a longer electrical resonator into a smaller unit cell [12], thus offering very appealing miniaturization skills. Furthermore, the fixed length of the radiating sides (i.e., the upper and lower sides of the patch in Figure 1(a)) guarantees a higher independence to mutual coupling effects. As a matter of fact, in [22], it is demonstrated how the stronger contribution to the mutual coupling between microstrip patches is that occurring along the -plane (-plane in Figure 1(b)).
(a)
(b)
2.1. Unit Cell Design
The layout depicted in Figure 1 is adopted to design a dual-band unit cell operating at the following 5G frequencies: and . A benzocyclobutene (BCB) polymer is adopted as a substrate material. The excellent material features in terms of low loss (), low dielectric constant (≅2.65), and strong dielectric stability against frequency and temperature make the BCB polymer a good choice for designing mmw microstrip antennas [23, 24]. A substrate thickness equal to 0.26 mm is fixed [24]. A commercial full-wave code [25], based on the infinite array approach, is adopted for unit cell analysis. A normal incident plane wave is considered (). A periodicity of 4.3 mm is fixed in both directions (i.e., ) corresponding to 0.4λ at 28 GHz and 0.54λ at 38 GHz. Following the design rules outlined in [12], the unit cell is designed to achieve the desired dual-resonant behavior. In particular, remembering that the use of smaller patch lengths gives higher resonant frequencies and, conversely, the use of greater values allows to move down the resonant frequencies, the size of the fractal patches is fixed to the values depicted in Figure 2. The minimum distance between two adjacent elements (i.e., , see Figure 1) is equal to 0.22 mm. Both the phase (Figure 2(a)) and the amplitude (Figure 2(b)) of the unit cell reflection coefficient are illustrated in Figure 2. Similar phase curve behavior vs. frequency can be observed at both resonances (Figure 2(a)). Furthermore, low losses smaller than 0.7 dB can be appreciated in Figure 2(b).
(a)
(b)
2.2. Parametric Analysis of the Unit Cell
The reflection phase tuning is independently realized at each resonant frequency , by varying the corresponding fractal scaling factor , leaving unchanged the patch size . Both factors, S1 and S2, are properly tuned to achieve a full-phase tuning range, at the corresponding operating frequency. Figure 3 shows the effectiveness of the adopted phase-tuning mechanism that allows to independently vary the reflection phase at each desired frequency. In particular, it can be observed that, by changing the scaling factor associated to the 28 GHz resonant element and leaving unchanged the other fractal sizes, the phase response of the unit cell varies only corresponding to a neighborhood of , whilst a negligible phase variation is obtained at (Figure 3(a)). A similar behavior can be observed in Figure 3(b). In fact, by varying the scaling factor , a variable phase shift is introduced only corresponding to a neighborhood of , leaving unchanged the phase response at . The phase curves computed vs. the scaling factors (Figure 4) show the admissible scaling factor variation range for both S1 (Figure 4(a)) and S2 (Figure 4(b)). In other words, the results depicted in Figure 4 affirm that, in order to achieve a quite full-phase tuning range (≅320°) at 28 GHz, the scaling factor must be varied from 0.31 up to 0.45, assuring also small phase excursions, corresponding to the frequency , that are comprised between the phase error margins given by (Figure 4(a)). Conversely, the scaling factor must be varied within the value range starting from 0 up to 0.29, in order to assure a quite full-phase range at 38 GHz and a confined phase excursion error corresponding to the frequency (Figure 4(b)).
(a)
(b)
(a)
(b)
In order to give a more exhaustive description of the reflection phase response computed for the designed dual-band unit cell, a contour plot is reported under Figures 5 and 6, showing, respectively, the phase variations computed at frequency vs. , for different values (Figure 5), and the phase computed at frequency vs. , for different values (Figure 6). Both figures confirm how the proposed dual-band unit cell can offer an independent phase-tuning mechanism at each designed frequency band. As a matter of fact, a quite constant reflection phase can be observed at 28 GHz (Figure 5), by changing the scaling factor for a fixed value. Similar considerations can be extrapolated from Figure 6.
The unit cell is further analyzed to investigate the behavior of its phase response, for different incidence angles of the impinging wave. Figure 7 shows acceptable phase variations under 30° and 40° oblique incidence angles, with respect to the normal case. Furthermore, as depicted in Figures 7(b) and 7(c), the independency of the phase-tuning mechanism is preserved at both frequencies also in the case of oblique incidence angles. In particular, it can be observed that, by changing the scaling factor associated to the 28 GHz resonant element, negligible phase variations are obtained at the frequency (Figure 7(a)). A similar behavior can be observed in Figure 7(b), in the case of the 38 GHz resonant patch.
(a)
(b)
(c)
3. Dual-Band Reflectarray Designs
In order to prove the effectiveness of the proposed dual-band unit cell, two small reflectarray designs are reported in the following. The high independence between the two designed frequency bands (28/38 GHz), demonstrated in Section 2.2, allows to separately synthesize the two sets of resonant fractal patches (identified by and in Figure 1), without affecting the accuracy of the adopted design procedure. Both reflectarrays are designed by adopting a synthesis algorithm [9] that receives as input the desired radiation pattern, in terms of the main beam direction and maximum side lobe level, and automatically returns the required excitation phase on each reflectarray cell. The algorithm uses the phase design curves depicted in Figures 4(a) and 4(b) that assume fixed fractal patch sizes, except the adopted phase-tuning parameter (i.e., for and for ). A normal incident plane wave is considered.
The first design consists of a reflectarray able to independently scan the main beam at the two different operating frequencies, whilst the second design is characterized by a multibeam pattern.
The fractal patches embedded in each cell of the first reflectarray are chosen as follows: the elements resonating within the 28 GHz band, identified by the couple (, ), are computed giving the main lobe steered towards the direction , in the -plane (-plane in Figure 8); whilst the elements resonating within the 38 GHz band, identified by (, ), are chosen by imposing a main beam direction equal to -20°, in the -plane. Figure 8 shows the synthesized reflectarray layout (Figure 8(a)) and the full-wave simulations of the overall structure. In particular, two independent radiation patterns are achieved at the two designed frequencies. Both patterns fully satisfy the constraints imposed during the synthesis stage, confirming the effectiveness of the proposed dual-band unit cell.
(a)
(b)
Furthermore, a dual-band reflectarray is designed to achieve a multibeam radiation pattern corresponding to the first resonance (i.e., 28 GHz). In particular, a pattern having three beams, respectively, directed along -30°, 0°, and 30° in the -plane, is synthesized at 28 GHz, whilst a broadside radiation pattern is imposed at 38 GHz. The full-wave simulation of the overall reflectarray structure is depicted in Figure 9. The simulated patterns match quite well the constraints imposed during the synthesis stage.
Finally, a reflectarray is designed to independently direct the main lobes at the two different operating frequencies in the -plane (i.e., at 28 GHz and at 38 GHz). The reflecting surface is illuminated by a Ka-band horn (characterized by a 15 dB gain and a aperture), which is placed in the -plane (i.e., the -plane in Figure 10), at a distance of 16.5 cm from the reflecting surface, with an offset angle of about 15°. The two couples of elements are properly synthesized to compensate the phase delay in the paths from the feed and to introduce a proper phase contribution able to meet the synthesis constraints. Figures 11 and 12 show the contour plot of the synthesized patterns in - coordinates, where and . Figure 13 illustrates the gain patterns vs. frequency, computed, respectively, along the main beam direction , in the case of the 28 GHz radiation pattern, and along the direction , in the case of the 38 GHz radiation pattern. A greater gain peak value can be observed in the case of the 38 GHz pattern (i.e., the gain difference is about 4.2 dB), mainly due to the greater electrical size and the lower spillover characterizing the antenna aperture at the higher operating frequency (38 GHz) and secondly due to lower scan losses [26]. Although the structures have not been properly optimized in terms of bandwidth, the simulated gain patterns show an acceptable 1 dB gain-bandwidth, approximately equal to 950 MHz at 28 GHz (≅3.4%) and to 1.5 GHz at 38 GHz (≅3.9%), that, at the considered mmw frequencies, can potentially offer high data rates [1], as required by 5G applications.
(a)
(b)
The above reflectarray designs confirm the effectiveness of the proposed dual-band unit cell in achieving an independent control of the reflectarray radiation features at each operating frequency. As future developments, the proposed configuration will be further optimized in terms of bandwidth. Furthermore, experimental validations of the configuration will be scheduled in order to deal with the problems associated with the manufacturing process, such as fabrication tolerances and illumination feed displacement, not yet considered in this work.
4. Experimental Validation of the Unit Cell
In order to give a preliminary validation of the proposed dual-band unit cell, a small array prototype, composed of identical cells, is realized and tested in the Microwave Laboratory of the University of Calabria (Figure 14). The array is printed on a DiClad 880 substrate, having and [27]. A periodicity equal to 5 mm is fixed in both directions. The patches embedded in each unit cell are characterized by the following dimensions: , and , , giving a minimum distance between two adjacent patches equal to about 0.33 mm. The cell operates in a dual-band mode corresponding to about 29 GHz and 39 GHz.
A far-field measurement system is adopted (Figure 14), consisting of two identical transmitting and receiving horn antennas (operating within the () GHz frequency band), both connected to a vector network analyzer. The instrumentation detects the field reflected by the array along the broadside direction in the far-field region [9]. The reflection phase curve of the cell is measured within the frequency range (Figure 15). It can be observed a phase variation of about 300° around both operating frequencies. Furthermore, a good agreement between simulations and measurements can be observed in Figure 15, showing a small but appreciable frequency shift, mainly due to the fabrication error tolerance related to the adopted printed circuit board (PCB) milling process [28]. Anyway, the effects due to the above errors can be reduced through a proper reiteration of the unit cell synthesis procedure (see Section 2) or, alternatively, by adopting a more precise fabrication process, as that described in [23, 24]. In conclusion, it can be stated that the experimental results confirm the dual-band behavior of the proposed configuration.
5. Conclusion
A single-layer dual-band reflectarray with a single linear polarization cell has been investigated in this work for future 5G wireless communication networks. A reflectarray unit cell operating within the Ka band (at 28/38 GHz) has been designed. Two pairs of miniaturized fixed-length fractal patches have been synthesized, achieving low losses (<0.7 dB) and almost full-phase ranges (≅320°), at both operating frequencies. A thorough parametric analysis of the unit cell has been performed, demonstrating negligible mutual coupling effects between the two pairs of resonant elements, so assuring the independence between the two designed frequency bands. The designed compact cell has been successfully adopted to demonstrate reflectarrays’ abilities in achieving fixed scanned-beam and/or multibeam patterns, under the dual-band operation mode.
Full-wave simulations of the synthesized reflectarray structures have been performed, confirming the effectiveness of the designed dual-band configuration in achieving independent radiation patterns at the two designed frequencies. In conclusion, the designed dual-band configuration offers, at the same time, compactness, low losses, frequency diversity, high versatility in achieving fixed scanned-beam and/or multibeam radiation patterns, and quite good gain-bandwidths. The above features make the proposed dual-band reflectarray configuration appealing for the implementation of 5G antennas. A preliminary experimental validation of the dual-band behavior of the proposed unit cell has been performed.
As future developments, the proposed configuration will be further optimized in terms of bandwidth and fabrication tolerances. Further experimental validations on dual-band reflectarray prototypes will be scheduled for the future, in order to give a comprehensive proof of the proposed concept.
Data Availability
The data used to support the findings of this study are included within the article.
Conflicts of Interest
The authors declare that they have no conflicts of interest.