Building Toward 6G: Arrays, AI, Spectrum Sharing, And The Next Measurement Frontier

By John Oncea, Editor

Wi-Fi 7 and 5G-Advanced mark a turning point in RF front-end design, where integration, AI-driven optimization, and spectrum-proof architectures are shaping 6G foundations.

The wireless industry is at a hinge point. Wi-Fi 7 equipment is entering the market following the Wi-Fi Alliance’s 2024 certification program announcement, which supports multi-link operation and 320 MHz channels for multi-gigabit throughput. Meanwhile, 3GPP Release 18, frozen in June 2024, marks the start of the 5G-Advanced phase and formalizes new work items across positioning, energy efficiency, and AI/ML interfaces for radio management.

As this transition unfolds, U.S. public research agendas have shifted toward spectrum sharing science, measurement traceability, and early 6G frameworks. The National Telecommunications and Information Administration (NTIA) published a request for input on 6G use cases and research priorities in mid-2024, framing 6G not simply as an extension of 5G but as a system that must integrate communications with sensing, security, and coexistence among commercial and federal users.

These programs and standardization developments converge on a shared set of RF design pressures. Front ends must cover more bands, steer more beams, consume less power, and maintain spectral cleanliness while operating within shared regulatory environments that continue to evolve.

mmWave And High-Frequency Front Ends

Wi-Fi 7 access points increasingly use multi-band, multi-chain transmitter, and receiver paths to maintain aggregate throughput and reliability. The extension of Wi-Fi into the 6 GHz band for unlicensed operations has raised coexistence stakes, requiring cleaner power amplifier operation, sharper filtering, and lower-loss transitions between front-end components and antennas.

On the cellular side, the mmWave portion of 5G New Radio remains a premium performance layer, but Release 18 adds features that strengthen the link budget and improve usability in more deployments. Power amplifier efficiency and antenna-in-package integration continue to be central constraints.

Supporting these developments, the National Institute of Standards and Technology (NIST) has published ongoing research and calibration guidelines at mmWave frequencies to enable traceable measurement environments from 26.5 GHz through 110 GHz. Such metrology enables RF engineers to evaluate true system-level performance rather than relying solely on isolated device characterization.

Massive MIMO And Adaptive Beamforming

Modern base stations, fixed wireless systems, and dense enterprise access points increasingly rely on multi-element phased arrays. Dozens to hundreds of radiating elements require phase-coherent distribution networks and calibration loops capable of compensating for temperature and frequency drift. These effects directly impact beam quality, sidelobe suppression, and link stability under mobility and blockage.

Release 18 introduces improved positioning reference signals and signal structures that refine angle-of-arrival and time-difference-of-arrival estimates, tightening the performance requirements for phase noise and group delay consistency across array elements, according to 3GPP. As networks densify and operate with more spatial reuse, array-level thermal behavior and feed network parasitics matter as much as device specifications.

Reconfigurable And Agile RF Front Ends

A second major shift is the move from fixed-function RF front ends to architectures capable of reconfiguration in real time. With spectrum access spanning 2.4 GHz through 7 GHz for Wi-Fi and sub-6 GHz plus mmWave for cellular, multi-band hardware flexibility reduces device complexity and allows faster adaptation to surrounding interference conditions.

This adaptivity aligns with Release 18’s formalization of AI/ML-assisted radio control interfaces, which define how models can access telemetry and push configuration updates to the PHY and RF layers, according to 3GPP. Instead of simple static band-selection, next-generation front ends will participate in closed feedback loops where RF behavior responds dynamically to user density, channel quality, and application demand.

Power Efficiency And Thermal Management

The rise of antenna-in-package and system-in-package integration increases thermal density. Maintaining efficiency under elevated junction temperature is increasingly difficult at mmWave, where device efficiency is already lower than in sub-6 GHz power transistor technologies. This constraint applies across access points, smartphones, customer premises equipment, and fixed wireless transceivers.

The U.S. Department of Energy and NIST have collaborated on research efforts to define power-aware RF device and system characterization conditions, emphasizing that link-level efficiency must be considered rather than per-device PAE figures of merit, according to NIST.

AI/ML For RF Optimization

AI/ML is now positioned to influence beam management, predictive interference avoidance, and adaptive filtering. But the performance of AI-driven RF control depends heavily on accurate channel and spectrum models. Poorly calibrated models lead to unreliable or non-generalizable behavior. To address this, the NIST NextG Channel Model Alliance provides validated, shared channel measurement resources explicitly intended to improve the robustness of optimization algorithms.

These shared channel models allow RF behavior predictions to match observed propagation in dense and dynamic environments, improving beam selection and link adaptation outcomes.

Sub-THz And D-Band Research

Research activity is expanding in the 100 GHz to 300 GHz range to support exceedingly high data rate backhaul and high-resolution sensing. Device-level challenges remain substantial: transistor cutoff frequencies, packaging and interconnect losses, and free-space path loss all require phased arrays and beam concentration to achieve usable link budgets.

The Next G Alliance published a technical report on sub-THz feasibility and system considerations that outlines integration challenges, propagation characteristics, and packaging constraints in the 220–330 GHz D-band region. Commercial adoption is still emerging, but the findings from sub-THz research are already influencing packaging and assembly strategies in the upper mmWave regime.

Integrated Sensing And Communications (ISAC)

ISAC has gained significant traction as a 6G research priority. Rather than deploying separate systems for radar and communications, ISAC envisions shared waveform designs and shared antenna apertures. This integration reduces hardware duplication and enables sensing functions for positioning, gesture detection, channel-aware signal adaptation, and environmental awareness.

The Next G Alliance published a foundational ISAC report outlining channel measurements, sensing use cases, waveform compatibility considerations, and phase-coherence requirements, according to the Next G Alliance. These findings imply that RF chains must support higher dynamic range, better oscillator stability, and calibration schemes that preserve timing and amplitude integrity.

Spectrum Sharing And Coexistence

Shared and flexible spectrum is essential for both Wi-Fi and 5G. The NTIA’s 2024 inquiry emphasizes that 6G must ensure interoperability among commercial services, satellite links, scientific use cases, and federal systems. This includes sharing frameworks, emission mask compliance, and real-time coordination.

For RF front ends, the implication is clear: spectral purity and out-of-band suppression must improve even as power density and complexity increase.

Preparing For The 6G Decade

The progression from Wi-Fi 7 and 5G-Advanced toward early 6G concepts is defined not only by higher frequencies and larger arrays but by increased interdependence between RF hardware, baseband processing, and adaptive control. Traceable measurements and realistic channel modeling are becoming as important as device-level improvements.

The RF design community now faces a landscape where integration, adaptability, and verification are inseparable. Teams that treat packaging, thermal design, calibration, and AI-assisted optimization as part of a coordinated engineering process will be positioned to deliver systems that scale from 6 GHz through mmWave and into sub-THz.

The transition is already underway. The work now is to turn these measurement-anchored capabilities into modules, arrays, and systems that can perform reliably in real deployment environments.