Enormous Fluid Antenna System Passes First Real-World Test, Guiding Radio Waves Across Programmable Surfaces

Researchers have unveiled and experimentally validated the first Enormous Fluid Antenna System (E-FAS), a radical new architecture that could redefine how future 6G networks move radio signals. Instead of relying solely on fragile over-the-air propagation, the system guides electromagnetic waves along a reconfigurable surface—like a programmable superhighway for millimeter-wave signals—and then launches them into free space with high-gain, steerable beams. The work, published in Frontiers in Communications and Networks, represents the first experimental demonstration of this hybrid guided-wave and radiating platform, and it tackles one of the most stubborn problems in next-generation wireless: high-frequency signals that attenuate rapidly, bend poorly around obstacles, and lose coverage over distance.

The core insight behind E-FAS is elegantly simple. Conventional wireless communication at millimeter-wave frequencies depends on radio waves bouncing through free space, suffering substantial propagation loss with every reflection and every meter traveled. The E-FAS prototype replaces a significant portion of that over-the-air journey with surface wave propagation—electromagnetic waves that cling tightly to a dielectric-clad conducting ground plane and travel along it with remarkably low attenuation. The measured insertion loss is approximately 0.1 dB per meter, a dramatic improvement over free-space propagation at these frequencies, where signals degrade far more quickly and coverage becomes severely limited.

The physical realization of this concept is built on a porous, dielectric-coated metal ground plane featuring a dense array of hollow cylindrical cavities spaced at sub-wavelength intervals of just 1.2 millimeters. Each cavity can be switched between two states: an air-filled off-state and a metal-filled on-state, achieved either by physically inserting or removing a metal rod of matching radius and length, or more practically by connecting electronic switching devices such as PIN diodes between the metal rods and the ground plane. When a control unit—an FPGA or microcontroller—forward-biases the PIN diodes, the rods connect electrically to the ground, creating a guided structure. Isolating slots separate the ground and the rods to achieve the off-state. Switching speeds range from a few microseconds down to nanoseconds, depending on the control modules and semiconductor devices used.

The physics underlying the system relies on Transverse Magnetic surface waves guided along the dielectric-clad conducting platform. The propagation constant of these waves is fundamentally determined by the surface reactance of the dielectric layer, which in turn depends on the free-space wavenumber and the intrinsic impedance of the medium. The researchers selected the F4BM-2 substrate for its stable dielectric constant of 2.2 and low dissipation factor of 0.001 at millimeter-wave frequencies, allowing the platform to operate across temperatures from −50 °C to +260 °C with minimal moisture absorption—less than 0.09 percent change when dipped in 20 °C distilled water for 24 hours. These material properties ensure consistent electromagnetic performance under real-world environmental conditions.

By dynamically manipulating the operating states of the cavities, the design establishes highly efficient surface wave channels that steer electromagnetic waves across the tailored reactive surface. The surface-wave channel is formed by configuring two rows of cavities in the on-state, and the width of this channel determines the operating frequency—much like a conventional metal waveguide. By adjusting the channel pathway, the signal can be routed to different positions on the surface. Simulated electric-field distributions confirm that the surface wave generated by the feeder remains confined within the defined path, achieving high-efficiency isolated propagation without significant leakage into free space during transit.

Crucially, the architecture facilitates reconfigurable surface wave launchers—specific distributions of on-state cavities—that radiate high-gain beams into free space at precise locations and directions. The launcher is composed of two ±45° lines of 1 × 8 on-state cavities and a 7 × 10 rectangular array of on-state cavities. In this configuration, the on-state cavities act as scatterers that gradually leak energy as the surface wave propagates through them, functioning like a leaky-wave antenna. Simulated radiation patterns show a directional beam with gain of about 19 dBi at 30 GHz, with a maximum at 30° above the horizontal plane. By strategically arranging different cavity patterns, the surface wave can be transformed into a single focused beam at a specific elevation angle or simultaneously divided into multiple distinct beams—all without any physical alteration of the hardware.

To validate the concept experimentally, the team fabricated a prototype consisting of a 20-centimeter-long surface wave channel integrated with a surface wave launcher, for a total length of 25 centimeters (25 wavelengths at the operating frequency) and a width of 5 centimeters. During measurement, electromagnetic waves radiated from the E-FAS were collimated by a reflector to produce a quasi-plane wave, which was then received by a standard-gain horn in a compact range antenna chamber. A laser positioning system aligned the antennas and reflector with high precision. The operating frequency range spans from 25.4 to 27.4 GHz, with the bandwidth primarily limited by the feeder design rather than the surface itself.

The measured results show reasonably good agreement between simulation and experiment. In the downlink scenario—where signals propagate from the E-FAS into free space—the maximum realized gain is about 18.5 dBi at 26.4 GHz within the passband. Deviations in the radiation patterns are likely due to fabrication and assembly tolerances, while oscillations visible outside the main beam direction arise primarily from phase interference between two parasitic radiation sources: direct radiation from the feeder into free space and leakage of surface wave energy during propagation. In the uplink scenario—where the launcher captures a wirelessly transmitted wave and converts it into a surface wave—the radiation level is approximately 2 to 4 dB lower across the frequency range of interest, attributed to the limited footprint of the launcher, which cannot capture all of the power radiated from the source horn.

These experimental results demonstrate that the Enormous Fluid Antenna System effectively bridges the gap between guided wave propagation and reconfigurable radiation, providing a low-loss alternative to traditional over-the-air links. However, the transition from laboratory prototype to commercial 6G deployment presents significant scalability challenges. The current prototype relies on precision machining of a porous surface with a high-density array of cavities and manual metal-rod insertion—suitable for validation but impractical for mass production. Large-scale implementation would require advanced PCB fabrication processes such as high-aspect-ratio laser drilling and automated surface-mount technology for integrating switching elements like PIN diodes across thousands of cavities. The researchers point to a particularly promising manufacturing path: the modular assembly methods established for large-scale LED display panels, which share the same fundamental geometry of a massive planar aperture composed of discrete, repeating sub-wavelength units. By adopting a tiled modular architecture, individual E-FAS panels could be manufactured using standardized automated processes and interconnected to form an effectively enormous radiating surface, with modular control units managing cavity switching states and ensuring precise timing and phase synchronization between adjacent panels.

From a cost perspective, the choice of high-performance dielectric substrates and the density of control circuitry are the main drivers of system expenditure, yet E-FAS may offer a superior cost-to-performance ratio compared to conventional fully digital massive MIMO architectures by significantly reducing the number of active RF chains required. Because reconfiguration is achieved through spatial switching of the radiating aperture rather than through independent phase-controlled active elements, a smaller number of shared RF sources will not incur the significant phase noise typical of massive MIMO. The hardware complexity is shifted from expensive RF transceivers to lower-cost passive radiating elements and control logic. Integration strategies for commercial 6G networks must also address the significant overhead associated with channel state information acquisition and real-time optimization in systems with massive numbers of reconfigurable elements. The research group is prioritizing the impact of uplink/downlink differences, scalability considerations, electronic control of cavity on/off states, and surface-wave launcher design as key aspects for future investigation, welcoming other researchers to join this challenging journey toward practical 6G and beyond wireless systems.

Subject of Research: First experimental demonstration of an Enormous Fluid Antenna System integrating low-loss surface wave guiding and reconfigurable high-gain beam radiation for 6G wireless communications

Article Title: Experimental demonstration of enormous fluid antenna system for future intelligent wireless systems

Article References: Liu, Q., Wu, F., & Tong, K.-F. (2026). Experimental demonstration of enormous fluid antenna system for future intelligent wireless systems. Frontiers in Communications and Networks, 7, Article 1773573. https://doi.org/10.3389/frcmn.2026.1773573

Image Credits: AI Generated

DOI: 10.3389/frcmn.2026.1773573

Keywords: fluid antenna system, 6G, surface wave communication, reconfigurable surface, millimeter wave, PIN diodes, leaky-wave antenna, beamforming, massive MIMO, wireless propagation, reconfigurable intelligent surface, Frontiers in Communications and Networks