$$\rightleftharpoonup{xx}$$
$$\longleftharp{xx}$$,
$$\longrightharp{xx}$$,
Antenna design and simulation setup
The antenna was designed using commercial 3D electromagnetic simulation software based on the finite integration technique (FIT)10. The simulation parameters were configured as follows: (i) solver type: time domain solver with hexahedral mesh; (ii) frequency range: 4–7 GHz; (iii) mesh density: 20 lines per wavelength, adaptive mesh refinement enabled with accuracy threshold of –30 dB; (iv) boundary conditions: open (add space) in all directions with minimum distance of λ/4 from the structure; (v) farfield monitors: configured at 0.1 GHz intervals across the frequency range; (vi) port excitation: waveguide port with 50 Ω impedance. A Rogers RT/duroid 8550 was selected with relative permittivity (εr) of 2.55, thickness of 1.5 mm, and loss tangent (tan δ) of 0.0033. Visual checkpoint: Upon completion of the model geometry, the antenna structure should appear as shown in Figure 1, with clearly defined CRLH unit cells, interdigital capacitors, and T-stub inductors.
CRLH-TL antenna fabrication
The antenna employed an asymmetric coplanar waveguide (CPW) feeding structure. The series capacitance was realized using an interdigital capacitor, and the shunt inductance was implemented using a T-stub inductor, with all geometric dimensions defined according to the layout shown in Figure 2A. The shunt inductance was implemented via a T‑stub inductor (stem width: 0.5 mm, stem length: 15 mm, stub width: 6.0 mm, stub length: 6.0 mm). Instead of conventional via holes, a third‑order Hilbert curve structure (outer dimensions: 7.87 mm × 7.87 mm; trace width: 0.48 mm; gap: 1.19 mm; additional segment lengths of 2.39 mm, 1.43 mm, and 3.10 mm as shown in Figure 2B) was incorporated on the reverse side of the substrate to provide inductive loading while minimizing radiation losses11. Seventeen-unit cells were arranged in a 1 × 17 linear array with inter‑cell spacing of 8.5 mm (approximately λg/4 at 5.6 GHz). The interdigital capacitor and T-stub inductor dimensions were parametrically optimized; the final design corresponds to the configuration that yields maximum gain and minimum S11, as shown in Figure 2.
EBG structure implementation
A 17 × 1 array of third‑order Hilbert curve structures was printed on the back surface of the substrate, aligned concentrically with each T‑stub inductor. Each Hilbert element occupied an area of 7.87 mm × 7.87 mm, with the T‑stub inductor positioned centrally within this area to maximize inductive coupling12. The EBG array was positioned such that its longitudinal axis coincided with the centerline of the CRLH-TL array, with registration achieved using optical alignment marks during photolithography. The Hilbert EBG suppresses surface wave propagation and improves impedance matching13. Visual checkpoint: The third-order Hilbert fractal geometry used for EBG implementation is shown in Figure 2B.
AMC reflector design
A 7 × 10 AMC reflector array was designed using hexagonal unit cells (side length: 5.2 mm; periodicity: 9.0 mm). Each unit cell featured three rectangular slots (length: 3.5 mm; width: 0.4 mm) arranged at 120° intervals on the patch center to achieve zero reflection phase at 5.6 GHz. The array was fabricated on an identical Rogers RT/duroid substrate and positioned 8.0 mm (approximately λ0/6.7 at 5.6 GHz) below the antenna ground plane. Element spacing was maintained at 0.5 mm between adjacent hexagons, yielding a total array footprint of 200 mm × 100 mm. Alignment of the AMC array relative to the antenna was achieved using mechanical fixturing with micrometer‑adjustable stages, ensuring ±0.1 mm positional accuracy. Visual checkpoint: The AMC reflector unit-cell geometry and the complete array layout should correspond to Figure 3.
Optical switch integration
Surface‑mount light‑dependent resistors (LDRs) were integrated into the T‑section of each CRLH unit cell. Each LDR was positioned across a 0.5 mm gap in the T‑stub inductor stem, 3.0 mm from the stub junction. LDR attachment was performed using conductive epoxy cured at 80°C for 30 minutes. An external 635 nm red laser diode (optical power: 5 mW; beam diameter: 3 mm) served as the optical source. The laser beam was directed via a computer‑controlled galvanometer mirror system, enabling selective illumination of individual LDRs or groups of LDRs. For OFF‑state characterization, illumination was blocked using opaque black polyvinyl chloride (PVC) covers (thickness: 1.5 mm; dimensions: 10 mm × 15 mm) affixed over each LDR. Visual checkpoint: The LDR integration points within the T stub structure should be visible, as shown in Figure 1.
Fabrication and measurement process
Antenna fabrication employed a wet chemical etching process. Substrate panels were cleaned with isopropyl alcohol, dried at 60°C for 10 minutes, and laminated with positive photoresist (AZ 4562, spin‑coated at 3000 rpm for 30 seconds, soft‑baked at 100°C for 90 seconds)14. UV exposure (350 nm, 12 mW/cm2, 25 seconds) was performed through a photomask, followed by development in AZ 400K developer (1:4 dilution with deionized water, 45 seconds)15. Copper etching was conducted using ferric chloride solution (FeCl₃, 40% w/v, 40°C, 8 minutes)16. Residual photoresist was removed with acetone, followed by rinsing with deionized water and nitrogen drying17. Visual checkpoint: The fabricated antenna prototype as shown in Figure 4.
Measurements were conducted inside an RF-shielded chamber (dimensions: 4.5 m × 3.0 m × 2.5 m; isolation > 90 dB from 1–12 GHz). The antenna under test (AUT) was mounted on a computer‑controlled azimuth/elevation positioner (rotational accuracy: ±0.1°; angular range: 360° azimuth, ±90° elevation). A wideband dual‑ridge horn antenna (1–12 GHz) served as the reference antenna. S-parameter measurements were performed using a vector network analyzer calibrated with a full two-port SOLT (short-open-load-thru) calibration kit. Far‑field gain was determined using the gain‑transfer method with a standard gain horn of known gain (8.2 dBi at 5.6 GHz). All measurements were conducted under ambient laboratory conditions (temperature: 22 ± 1 °C; relative humidity: 45 ± 5%).