Low-profile flexible frequency-reconfigurable millimetre-wave antenna for 5G applications

The antenna design integrates frequency-reconfiguration and MIMO configuration on a flexible substrate which adds an advantage of antenna placement even on irregular surfaces. The CST STUDIO SUITE has been used for the analytical design and numerical modelling, which has been conducted based on the finite integration technique using the robust transient solver to evaluate the time-domain characteristics of the antenna structure. The design process is initiated with a single antenna element, comprised of a slotted T-shaped patch and a coplanar waveguide (CPW) feeding. This type of slotted patch geometry enables a compact and small footprint along with the incorporated switches for the reconfiguration [28]. The ground plane is designed with an etched-off rectangular aperture for the placement of the antenna.

Integration of a pair of symmetrical slots terminated with x-pair of switches is carried out on the patch to reconfigure the frequency based on the state of the switches. The flexible antenna resonates at 28 GHz when both the switches are OFF, as the slot-edges act as an open-circuit. While the resonant frequency band shifts to 38 GHz when both switches are ON, the radiating length is trimmed off as the slot-edges act as a short circuit. Furthermore, the design is modified with another pair of symmetrical slots provided with y-pair of switches which results in two additional operating modes. In order to diversify the flexible antenna functionality as a standalone transmitter and a receiver, a two-element MIMO assembly is introduced for the designed antenna configuration. Figure 1(a) shows the proposed MIMO antenna, and the dimensions of the design parameters are given in table 1. Moreover, DC-biasing is required to change the switch-state, while implementing the actual switches, e.g., PIN diodes. Here, if the x-/y-switch is applied with a potential difference, the metal short parallel to the switch, denoted by z in figure 1(a), would provide a path for the DC current, and the switch would not be enabled. Hence, a gap of 0.1 mm is inserted at z, along with a 0.1 μF capacitor, which blocks the DC from z without interrupting the AC flow and the DC current is directed through the switch.

Zoom In Zoom Out Reset image size

The designed MIMO antenna is fabricated on the surface-treated PET film from Mitsubishi Paper Mills Ltd, with a thickness (h) = 135 ± 12 μm, dielectric constant (ε_r) = 3.2, loss tangent (tan δ) = 0.002, and dimensions of 11 × 25.4 mm², as shown in figures 1(b) and (c). Moreover, the in-house Dimatix materials printer (DMP-2831) is used for the antenna prototyping, along with the silver nanoparticle ink (i.e., colloidal Ag-J solid Ag from Printed Electronics Ltd). The silver nanoparticles are encapsulated with the organic polymer coating to avoid oxidation, and are dispersed in an inert solvent. The selected ink comprises of 20.3% of solid silver content by weight, suspended in a solution of water (40%–50%), glycerol (30%–40%), and acetylene glycol (0.1%–1.0%). The conductive ink has a viscosity of 4.2 mPa s⁻¹, the surface tension of 30.5 mN m⁻¹, and density of 1.22 g cm⁻³. The post-printing processes of drying, curing, and sintering are essential to release the particles from the polymer sheath and combine them into a firm conductive layer.

The conductivity of the printed layer is 0.4–2.5 × 10⁷ S m⁻¹, and depends on the number of successive printing iterations, layer thickness, temperature, and time durations involved in both the curing and sintering processes. It is worth noting that an accurate and careful sintering results in the conductivity of 0.3–0.7 × 10⁷ S m⁻¹ from one printing iteration. To accomplish the fabrication, the printer has been appropriately calibrated in terms of the drop-spacing, jetting frequency, waveform, printhead temperature, and firing voltage of the ejecting nozzles. The drop-spacing of 15 μm (i.e., 1693.33 dpi), firing voltage of 15 V, and jetting frequency of 5 KHz have been set in this regard. The variation in the surface profile of PET is characterised by using a profilometer, as in figure 2(a), and utilised while adjusting the printhead position. A conductive layer with a thickness of 0.5 μm is deposited on a PET film with a pattern resolution of ±20 μm.

Zoom In Zoom Out Reset image size

The PET substrates are provided with a chemically active microporous adhesive coating to enable the rapid drying and chemical sintering of the ink. This chemical method avoids laser exposure or high-temperature treatments that may result in the bending, shrinking, decolourisation, or melting, if the temperature is not appropriately handled. The microporous coating elevates the surface energy of the PET and improves the adhesion by chemically bonding the ink with the surface molecules, and restricts the ink flow outside the boundaries of a definite pattern for a precise prototyping. Another advantage of the chemical coating is to replace the actual surface of PET with a microporous surface, which allows partial diffusion of silver nanoparticles to facilitate better locking between the ink particles and surface molecules that results in a high adhesion and prevents the layer from peeling off.

The conformal assembly of the antenna involves bending of the prototype, which may cause cracking of the layer with time. In this regard, the thickness of the printed layer is critical, as thicker metal layers are more brittle and can be easily cracked by successive bending and folding. For instance, the screen printed layer is usually 10–20 μm thick with relatively larger particle size, which makes it less flexible and more prone to cracks than the inkjet-printed nanoparticle layer (i.e., 0.5 μm thick). The antenna is firmly mounted on a cylindrical surface of a 0.8 mm diameter pen along the axial, circumferential, and diagonal axes, as in figure 2(b). The post-bending microscopic inspection depicts no cracks. In addition, the standard Scotch tape test is performed to examine the robustness [29, 30]. The tape is glued to the printed pattern and then peeled off while the printed trace remains fully intact on the substrate, as in figure 2(c), and exhibits the excellent ink adhesion and robustness. The prototype is also found water-resistant when submerged in water. Moreover, to obtain a clear insight, SEM imaging of the printed silver pattern is performed that shows the granular nature of the ink layer, as depicted in figures 2(d) and (e).

In order to inspect the antenna response toward conformity by the experimental testing and measurements, the fabricated MIMO antenna is mounted on K-connectors in both planar and conformal assemblies (curved surface of 3 cm radius). Figure 3 presents the measured results of the reflection coefficient (i.e., S₁₁/S₂₂) plots of the developed planar and conformal MIMO antenna for the proposed switch configurations, represented by the corresponding unit-cell and also shown in table 2. The overall bandwidth of 27.3–40 GHz has been covered in the four switching modes of the reconfigurable MIMO antenna.

The measured transmission coefficients in the operating modes are depicted in figure 4 that shows lower mutual coupling and high isolation between the two adjacent antennas mounted in a planar assembly. The standard distance between adjacent elements in an array should be in-between λ/2 to λ, to avoid mutual coupling, i.e., higher the distance lower the coupling; however, increasing the distance might result in grating lobes [31, 32]. The proposed antenna is a T-shaped patch with more width than a rectangular patch or λ/2, which suggests more spacing between the patches' centres. The optimal level of the isolation suggests the transmission coefficients of at least −10 dB or below. It has been observed in the parametric study for achieving a high isolation, the optimal distance between the centres should be approximately 11 mm (i.e., equals to λ at 27.2 GHz that is the lower cut-off frequency). However, the distance between the centres is slightly increased due to larger patch width and is set at 12.7 mm in this proposed MIMO antenna, to accomplish a high isolation below −20 dB. The envelope correlation coefficient (ρ_e) is computed to evaluate the independence of the antenna elements for their individual operation in a MIMO assembly, and calculated based on the S-parameters. In this regard, figure 5 presents the low values of the numerically estimated ρ_e of the MIMO antenna, which depicts insignificant influence of neighbouring element on the performance of the single antenna.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The radiation characteristics of the developed antenna are measured by using the NSI near-field scanning setup for K_a-band that exhibits a precise near-field to far-field algorithm to compute the far-field radiation patterns. An evenly distributed pattern on the front and bottom with the maximum directivity perpendicular to the surface of the antenna has been observed throughout the operating range. Figure 6 presents the simulated and measured results of a single element (i.e., a unit-cell of the MIMO) at φ = 0° and φ = 90° at distinct frequencies of the operating modes I–IV. As the patch antenna is placed in the aperture cut inside the ground, it causes the surrounding metallic ground to act as a loop or frame, which confines the radiation of the patch orthogonal to the antenna plane (i.e., along the front and back directions), while the radiation is partially trimmed off along the antenna plane, as shown by the two orthogonal cuts (i.e., E- and H-planes) of the figure 6.

Zoom In Zoom Out Reset image size

In a MIMO antenna, each element operates individually, hence, the peak gain of a single antenna is more important to be determined rather than the gain profile of a collective array. The simulated and measured results of the realised peak gain for a single antenna (i.e., unit-cell) of the proposed MIMO configuration over the operating range of 28–40 GHz have been tabulated in table 3 for the operating modes I–IV. The gain is computed based on the measurements carried out in the near-field scanning, and the two-antenna gain method is also utilised for the calculations. The measured gain shows similar findings as obtained by the simulated evaluations.

The proposed MIMO antenna, which depicts the advantage of combining the flexibility, reconfiguration mechanism, and MIMO technology within a single antenna prototype has been compared with the recent progress of 5G antenna designs in table 4. Several flexible and wearable antennas are reported for 5G, and significant work is also conducted at the mm-wave bands to develop MIMO antennas for 5G along with adaptable antennas for the frequency selection. In this regard, this work has undertaken the aggregation of these advanced wireless technologies, in order to deploy a novel and efficient antenna solution for the next-generation communication systems.