Video calls die in parking garages. Signals vanish in packed stadiums. 5G promised to fix this, and it largely didn’t. 6G can’t afford the same outcome, and engineers are finally treating coverage gaps as a geometry problem rather than a power problem.
Kongju National University filed a March 2026 Korean patent describing a four-plane 3D Reconfigurable Intelligent Surface, or RIS, a type of programmable antenna panel that can steer wireless signals without active amplification. It’s the most recent entry in the public 6G RIS dataset and a clear signal that next-generation hardware is moving past flat panels [Patsnap]. With 6G standardization accelerating through 2026, the choices made this year will shape what gets bolted onto lamp posts for the next decade.
Why 6G Coverage Demands a New Approach
6G research targets end-to-end latency below 1 ms, connectivity for 10 million devices per square kilometer, and full three-dimensional coverage [RSIS].
Photo by Those aren’t marketing slides. They’re the working assumptions inside ITU-R and ETSI, the international bodies drafting the technical standards.
Flat-panel RIS architectures can’t deliver that. Millimeter-wave and sub-terahertz signals, the spectrum 6G depends on, get absorbed by walls, foliage, and human bodies within tens of meters. A planar RIS mounted on a wall redirects energy across roughly one hemisphere. Anything behind it or steeply above it stays dark.
The user distribution has also changed. 6G networks must serve ground-level pedestrians, vehicles, indoor IoT sensors, and aerial drones at the same time. A single angular sector no longer covers where users actually are. They’re scattered across a full sphere around every access point.
From Flat Panels to 3D Geometries
The first key milestone was near-field beamforming.
Photo by Tsinghua University demonstrated RIS beamforming optimized for receivers inside the Fresnel zone, the near-field region where signal behavior differs from open-air propagation, in 2023. That proved metasurfaces could shape wavefronts with precision rather than just bouncing them [Patsnap]. Hardware followed: ESPCI Paris/CNRS showed PIN diode switching at 28.5 GHz, while the University of Surrey demonstrated varactor diode tuning at 3.5 GHz [Patsnap].
Those were all flat surfaces. The leap to three dimensions came from wrapping reconfigurable elements across multiple faces of a structure: cubes, four-plane prisms, cylinders. Each face covers an angular sector, and a central controller coordinates phase shifts across all of them.
Key traits of current 3D RIS prototypes:
-
Multi-face layout: 1-bit reflective arrays distributed across four or more planes
-
Near-passive elements: reflection rather than amplification, keeping power draw minimal
-
Fast reconfiguration: PIN and varactor diodes switching states in sub-microsecond windows
-
Omnidirectional DOA estimation: DOA, or direction-of-arrival, uses single-point random phase shifting for full-sphere signal acquisition [arXiv]
What the Benchmarks Actually Show
Current performance data is still largely simulation-driven, worth saying plainly.
Photo by Published prototypes and patent filings describe 3D RIS geometries that cover angular sectors flat panels physically can’t reach, particularly at high elevation angles where aerial users operate.
The practical implications break down into three areas:
- Coverage continuity: Multi-face structures address elevation and azimuth simultaneously, closing the overhead gap that blocks drone corridors and upper-floor windows.
- Interference isolation: Non-overlapping angular sectors let operators assign beams to users without the cross-talk that plagues dense flat-panel deployments.
- Energy efficiency: Because RIS elements reflect rather than transmit, a 3D structure adds coverage without the power cost of active MIMO arrays, which are traditional antenna systems that both send and receive signals.
The debate has shifted. It’s no longer whether 3D RIS helps. It’s which geometry, four-plane, cubic, or cylindrical, ships at scale first.
Where 3D RIS Gets Deployed First
Three deployment tiers are emerging, each with different economics.
Photo by Street furniture comes first. Lamp posts, traffic signal arms, and building corners are already wired for power and backhaul. They’re natural mounting points for 3D RIS nodes that need to cover pedestrians and vehicles at the same time.
Large indoor venues are the second tier. Airports, warehouses, and hospitals suffer persistent shadowing from equipment racks and moving crowds. Ceiling-mounted 3D RIS units can redirect signals around those obstructions without requiring a denser access-point grid.
The third tier is mobile and temporary. Deployable RIS units on vehicles or tethered drones offer pop-up coverage for emergency response, construction sites, and outdoor events where running fiber isn’t an option.
ETSI and 3GPP working groups are actively drafting RIS-related specifications through 2026. Operators who start evaluating geometries now will have input into how these deployments get standardized, rather than inheriting someone else’s defaults.
3D RIS isn’t speculative anymore. Patent filings through March 2026, working prototypes at 3.5 and 28.5 GHz, and near-field beamforming research have converged on a clear direction: 6G coverage will be shaped by volumetric metasurfaces, not flat ones . The remaining questions are about geometry, manufacturing cost, and standardization timing. For engineers and network planners watching the 6G roadmap, the practical move is to start benchmarking 3D RIS options against current flat-panel assumptions while the specification windows are still open.
