The core problem the team attacked is one of the oldest constraints in antenna theory. A conventional microstrip patch antenna element resonates when one of its dimensions is approximately half the free-space wavelength, and array elements are conventionally spaced at about half a wavelength apart for two reasons. First, that spacing prevents excessive mutual coupling between elements, which otherwise leads to power loss primarily through undesired surface wave interactions rippling across the substrate. Second, the spacing is critical for ensuring desirable radiation efficiency and pattern characteristics. Because both the elements and the gaps between them are tied to the wavelength, miniaturizing a patch array has historically meant accepting painful trade-offs in bandwidth, radiation efficiency, and gain.

The Florida group, led by Alexander Wilcher, Ariel David Cerpa, and Yong-Kyu Yoon, sidestepped those trade-offs with a defected ground structure built from point-symmetric, dimensionally optimized meander-line complementary split ring resonators, or ML-CSRRs, etched into the ground plane between closely spaced elements. A CSRR behaves electrically like a parallel LC resonator: at its resonance frequency, its impedance peaks and its reactance vanishes as the capacitive and inductive components cancel. When the terminal impedance of the resonator is much greater than the characteristic impedance of the line, the reflection coefficient approaches one, implying total reflection without the phase change that would lead to power loss. Engineered properly, the resonator rejects the surface waves that couple energy between neighbors, increasing isolation and allowing a miniaturized array with enhanced performance.

The numbers are striking. In the metamaterial array, or MMA, the 4-element unit cell has an element spacing of just 3 millimeters, approximately one-sixteenth of the free-space wavelength, while the reference microstrip patch arrays use a conventional spacing of 25.8 millimeters, about half a wavelength. Simulations using ANSYS High-Frequency Structure Simulator showed that as the element count grows, the area reduction reaches nearly 80 percent for an ideal simulated 64-element case. Fabricated and measured hardware confirmed reductions of 76.83 percent for the 4-element design and 77.02 percent for the 16-element design compared to their reference arrays, with realized gains of 6.22 dB and 9.38 dB respectively. In simulations of closely spaced arrays, the meander-line resonators delivered more than 3 dB of decoupling in diagonal coupling compared to identical arrays without them.

Placing the resonators point-symmetrically between elements was a deliberate design choice. Because coupling in the proposed unit cell is dominated by diagonally placed neighbors, owing to the locations of the feed pins, the symmetric configuration minimizes the number of ML-CSRRs required and provides a simple, repeatable design process. Each geometric parameter of the meander line tunes the resonance: lengthening the meander-line gap shifts the frequency lower, as do wider meander lines and more meander sections, while fewer meanders raise the frequency when total length and width are held constant. That flexibility means the structure can be tuned to essentially any frequency of interest, a property the team demonstrated by engineering the resonance to the 5.8 GHz target.

Cramming elements together creates a second problem beyond coupling: the radiated waves from closely spaced elements interfere and distort the pattern, since waves no longer constructively interfere at the array broadside without the conventional half-wavelength spacing. The team solved this with sequential phase feeding, driving elements with 0-degree and 180-degree offsets so that a broadside pattern is recovered despite the tiny pitch. The feed network combines quarter-wave transformers, equal power splitting, and phase control by offset electrical lengths, all designed for 50-ohm impedance. A bonus of this architecture is polarization reconfiguration: by changing the phase feeding scheme, a 2 by 2 MMA can switch polarization while retaining a radiation pattern and gain like a single patch, with the addition of beam steering capabilities.

Scalability was assessed across 4, 16, and 64 elements. Feed structures were first drafted in Keysight Advanced Design System and then transferred to HFSS for full-wave optimization. The results reveal a compelling trend: simulated MMAs with 2-to-the-N elements achieve similar half-power beamwidth and peak realized gain as a reference array with 2-to-the-(N-1) elements, meaning a compact MMA effectively behaves like a conventional array twice its linear size. Simulated MMAs also outperformed their reference counterparts in total efficiency before feedline integration, exhibited lower side-lobe levels, and showed broader 10 dB bandwidths. For the ideal 64-element case, area reduction approached 80 percent, and the team expects a crossover point in large arrays where greater gain can be obtained within reduced areas.

Fabrication was deliberately kept mundane, and that is part of the significance. After evaluating a low-loss Rogers 5,880 substrate, the team chose standard FR4 glass-reinforced epoxy laminate from Sunstone Circuits, accepting a much higher loss tangent of 0.02 versus 0.0009 and a dielectric constant of 4.4 versus 2.2. Calculations according to the Groiss model showed FR4 microstrip lines attenuate at 0.024 dB per millimeter, nearly five times the 0.005 dB per millimeter of Rogers material. A four-layer stack-up mitigated the damage: the second copper layer beneath the patches was left effectively empty to thicken the dielectric and improve bandwidth and radiation efficiency, the third layer served as the ground plane, and the fourth carried the feed lines close to the ground to confine fields and minimize spurious radiation. The 8 by 8 designs were not fabricated because excessive FR4 substrate losses and the complexity of compact feeding structures ruled them impractical, though simulations of the 64-element case indicated potential for further scaling with better materials.

Measurement validated the concept at two independent facilities. Near-field scans at the University of Florida used a calibrated probe rotated around the device under test as a scan table stepped through angles, with return loss data used to algorithmically approximate far-field characteristics. At Rohde and Schwarz in Dallas, Texas, a substitution method placed the transmitter and receiver beyond the Fraunhofer distance, with a stepper motor rotating the stage at 0.11-degree accuracy and a vector network analyzer calibrated by the short-open-load-through method recording insertion loss at 10-degree intervals. Optical inspection with a calibrated microscope revealed deviations of up to roughly 11 percent in the fabricated meander-line dimensions, but Monte Carlo simulations accounting for plus-or-minus 20 percent tolerances had predicted this, and the variations did not significantly degrade antenna performance. Measured results closely matched simulations from both facilities.

The implications reach well beyond the laboratory. For size, weight, and power constrained platforms such as satellites and drones, an antenna array that delivers the beam control of a much larger conventional array within a fifth of the footprint changes what fits aboard the vehicle. More elements within a fixed area promise improved signal-to-noise ratio and finer electronic beam steering, improving a system’s immunity to electromagnetic interference, whether intentional or otherwise. The team envisions the 2 by 2 MMA as a foundational building block from which larger arrays can be designed for specific applications, benefiting compact wireless RF systems in both the defense and commercial sectors, including radar and communication applications. With adjustable element spacing offering a trade between realized gain and area reduction, designers now hold a dial that conventional patch arrays never gave them.

Subject of Research: Scalable metamaterial antenna arrays with suppressed mutual coupling for compact platforms

Article Title: Scalable metamaterial antenna arrays with suppressed mutual coupling for compact platforms

Article References: Wilcher, A., David Cerpa, A., & Yoon, Y.-K. (2025). Scalable metamaterial antenna arrays with suppressed mutual coupling for compact platforms. Frontiers in Antennas and Propagation, 3, Article 1695439. https://doi.org/10.3389/fanpr.2025.1695439

Image Credits: AI Generated

DOI: 10.3389/fanpr.2025.1695439

Keywords: Scalable, metamaterial, antenna, arrays, suppressed, mutual, coupling, compact, platforms, scientific research