6G news and analysis for the week ending Friday, 19th June, 2026.
Hello again everyone. It’s a quieter week for news, but to make up for it I am writing a report on the 3GPP SA # 112 meeting that I will publish over the weekend. It’s along the lines of the write-up on the RAN TSG and has taken quite a few evenings to put together!
The message coming across this week is that AI-native 6G has moved from aspiration into the stated industry direction, even as the economics that will fund it stay stubbornly unclear. Ericsson provided the week’s top story: Its French R&D centre is leading INTENTION-6G, a project worth more than €12 million backed by the France 2030 plan, to embed AI into RAN control for energy efficiency and intent-based operation. Partners include Orange, BubbleRAN and CentraleSupélec.
Nokia added some framing to all of this. In a post-EuCNC blog, the company argued that AI-native design is no longer an optional overlay but the defining characteristic of 6G. It is a fair reading of the mood in Málaga, though it does restate a consensus that has already been well rehearsed. As a counter, Dell’Oro Group poured some cold water on all this enthusiam. Its latest report projects 6G RAN capex reaching half a trillion dollars by 2034, yet warns that cumulative 6G RAN revenue in its first six years will run 10 to 20 percent below the equivalent 5G period.
Elsewhere, the GSMA published highlights of its 6G terminals study, Poland’s Microamp secured €6.5 million in EU funding for an mmWave platform, and the University of Huddersfield (go Yorkshire!) won a grant for a 6G test chamber. We also cover Jio Platforms’ patent dialogue with the EPO, plus four new research papers.
Ericsson is the leading partner in INTENTION-6G, a French research project to design AI-powered 5G and 6G networks. The work is supported under the France 2030 plan and co-funded by Bpifrance. The consortium includes Orange, BubbleRAN and CentraleSupélec, with a budget above €12 million over four years. Its central aim is to cut Radio Access Network (RAN) energy consumption while meeting 5G/6G traffic demands. The project has reached its mid-development phase, allowing the first technology building blocks to be defined. Highlighting Ericsson’s contribution to the project at the at the VivaTech 2026 event in Paris this week, Christian Leon, Head of Ericsson Western Europe, said the centre is “laying today the foundations for the networks that will support tomorrow’s critical use cases.”
What’s new: INTENTION-6G centres on integrating AI into network control at the RAN level. It targets advanced automation of network functions, dynamic energy optimisation and intent-based networks that adjust performance automatically to demand.
Why this matters: Energy efficiency is now a hard design constraint for 6G, not an afterthought. The project shows European public money is moving into the AI-native RAN agenda through a public-private model that combines vendor, operator and academic partners. Ericsson’s French centre has already contributed to more than 180 mobile patents. Intent-based operation is one of the slower-maturing 6G capabilities, so early demonstrator work carries weight. The initiative also advances Europe’s stated digital sovereignty ambitions in next-generation networks.
Implications for 6G standards: AI-driven RAN control and energy efficiency are active Release 20/21 study areas. Demonstrator data from projects like this feeds the evidence base that operators and vendors leverage for their 3GPP negotiations.
View source material to read more
Nokia has published a blog by Didier Bourse, its principal standardisation lead, arguing that AI-native 6G is now the settled industry direction. The piece reflects on the EuCNC & 6G Summit 2026 in Málaga, which we covered in recent editions. Bourse frames AI not as an optional overlay but as a foundational element of future networks. He sets out three takeaways: building scalable AI-native networking foundations, context-aware performance, and coordinated intelligence across devices, edge and cloud. Bourse wrote that AI-native design “is not just an add-on feature of 6G” but its defining characteristic. He also notes that commercial 6G deployment remains roughly four years away (which most of us agree on, but it is always worth highlighting that we are looking at 2030 onwards).
What’s new: The blog distinguishes two architectural tracks. The first treats AI as application-layer traffic, such as distributed inference and edge AI services. The second treats AI as a foundational radio and system technology, such as neural receivers and deep schedulers. Bourse argues these correspond to different KPIs and research challenges.
Why this matters: As a Tier 1 vendor, Nokia’s framing carries weight, but it largely formalises a consensus already visible across the industry. TelecomTV’s DSP Leaders World Forum event concluded that AI-native still lacks a settled definition and that today’s networks are being retrofitted rather than redesigned. The useful contribution here is the explicit split between “AI for network” and “network for AI”, which clarifies where standards work and revenue opportunities diverge. Bourse also stresses interoperability through standards as the counterweight to vendor-specific AI. There is no new commitment in the piece, but the direction of Nokia’s interest is confirmed.
Implications for 6G standards: The call for AI to operate across vendors and domains points to standardised AI/ML frameworks and data interfaces. These are active topics in 3GPP and O-RAN, with the foundational elements expected in Release 21.
View source material to read more
Dell’Oro Group has published its latest 6G Advanced Research Report, projecting 6G RAN capex toward $0.5 trillion by 2034. The headline figures repeat the firm’s established evolutionary thesis, which we have tracked through its October 2025 report and its May analysis. The sharper new point concerns revenue. Cumulative 6G RAN revenue in the first six years of the cycle is projected at 10 to 20 percent below the comparable 5G period. The firm argues that operators now sit in a stronger network capacity position than they did during the 4G to 5G transition, which dampens the case for early spend. Stefan Pongratz, VP of RAN and Telecom Capex Research, said the likeliest scenario is “still that 6G will be another G, with 6G RAN capex expected to accelerate toward the end of the decade.”
What’s new: The lower-revenue comparison is the report’s most striking element. The firm also notes that momentum behind spectrum above 7 GHz continues to build, alongside the established Sub-7 GHz and cmWave baseline.
Why this matters: This actually sharpens the investment logic rather than changing it. Operators will justify 6G on lower cost per bit and platform modernisation, not on speculative new revenue. A first cycle that earns less than 5G did is a sobering counterpoint to the AI-native ambitions noted elsewhere this week. It reinforces the case that vendors must present clear efficiency arguments well before deployment. Cumulative 6G RAN investment between 2029 and 2034 is still projected at around half of total RAN capex for the period.
Implications for 6G standards: Tight economics raise the pressure on 3GPP to deliver efficiency gains in Release 21, not just new capabilities. The evolutionary, Massive MIMO-led baseline aligns with the standing consensus on reusing the existing site grid and Sub-7 GHz spectrum.
View source material to read more
The GSMA has released highlights of a 6G Terminals Study Report, the terminals white paper it flagged for launch around MWC Shanghai in its May progress report. The full report is registration-gated, so only highlights are currently available. It argues that terminals will evolve from communication devices into intelligent, collaborative and autonomous platforms. These would integrate communication, sensing, computing and AI-native functions in a single device. Use cases cited include immersive XR entertainment, AI agents, humanoid robots and smart manufacturing.
What’s new: The highlights list key capabilities for future terminals. These include AI-native intelligence and AI agent support, integrated communication and sensing, multi-dimensional perception, and collaborative edge and cloud computing. The report also frames new monetisation routes through AI-enabled services and immersive experiences.
Why this matters: Terminal capability shapes what networks must support, so this is a standards-adjacent input rather than handset speculation. The framing aligns with the broader AI-native direction and with 3GPP’s integrated sensing and AI work. The device-heavy emphasis, including humanoid robots, sits some distance from near-term deployment. The study sits alongside the 6G architecture work the GSMA detailed in its May Community Progress Report.
Implications for 6G standards: Terminal requirements feed ITU-R, 3GPP and GSMA activity on the 6G system definition. AI-native and integrated sensing capabilities in devices map directly to active Release 21 study areas.
View source material to read more
Poland’s Microamp has secured €6.5 million from the European Innovation Council (EIC) Accelerator under Horizon Europe. The backing comprises a €2.5 million grant and €4 million in equity from the EIC Fund. The funding supports its Any-G mmWave AI-RAN platform, described as a reconfigurable, software-defined system. Microamp says the platform adapts across 3GPP standards, frequency bands and network modes through software updates or swappable radio frequency modules. The company positions itself as a prospective European champion for sovereign, indigenous next-generation wireless infrastructure. CEO Dawid Kuchta said “mmWave will be the highway of the AI industrial revolution” and warned that without European technology, critical infrastructure would be “controlled and dominated by external incumbents.”
What’s new: The EIC equity component is notable, signalling EU appetite to back deep-tech hardware rather than grants alone. Microamp claims the platform can cut infrastructure ownership cost fivefold, a figure it has not independently substantiated.
Why this matters: European technological sovereignty in radio hardware is a live policy concern and EIC support reflects it. The software-defined, upgradeable model addresses a limitation of rigid legacy infrastructure. The funding is modest against incumbent R&D budgets, so the platform’s market impact is unproven. The company’s roots are in 5G mmWave, and its 6G claims are obviously forward-looking marketing rather than deployed capability.
Implications for 6G standards: A reconfigurable platform that tracks evolving 3GPP releases could ease migration for private and industrial networks. However, supporting future 6G features such as integrated sensing (ISAC) and non-terrestrial networks (NTN) remains a design aspiration here, not a standardised capability.
View source material to read more
The University of Huddersfield in the UK has secured a £400,000 grant from the EPSRC to fund an electromagnetic anechoic chamber for 6G research. The award, from UK Research and Innovation, covers a chamber with a spherical near-to-far field scanner. It will support antenna measurement and higher-frequency radio research for 5G/6G, satellite communications and advanced sensors. The successful bid was led by Professor Pavlos Lazaridis, who said the facility “will increase our research capabilities and place us at the top of 6G research.”
What’s new: The chamber follows an earlier EPSRC grant for a 110 GHz vector network analyser and related test equipment. The combined kit will support two Horizon Europe projects, ISAC-NEWTON and 6G-ICARUS.
Why this matters: Measurement infrastructure is foundational and often overlooked. Accurate antenna and propagation characterisation at higher frequencies is a prerequisite for the FR3 and sub-terahertz bands under study for 6G. ISAC-NEWTON, focused on integrated sensing and communications, involves 17 academic and industrial partners. 6G-ICARUS spans 22 organisations across Europe, the US and Brazil. UK academic capability in this area supports the wider European research base.
Implications for 6G standards: Real-world antenna and channel measurements feed the propagation models that underpin 3GPP and ITU-R work. Sensing research through ISAC-NEWTON aligns with active integrated sensing study items.
View source material to read more
The European Patent Office (EPO) held an online meeting last month with Jio Platforms, India’s largest mobile network operator (details of which have only just been published). The discussion covered patent quality, the European patent system and standard essential patents (SEPs). The Jio delegation was led by Pradeep Kumar Bhatnagar, who heads Jio Labs 5G/6G and AI product development. Jio highlighted a portfolio of more than 6,000 patents across India, the US, Europe and other jurisdictions. Both sides agreed to a technical meeting in late 2027 to discuss Jio’s European patent applications.
What’s new: Jio expressed particular interest in standards and SEPs, which it sees as central to its 5G activity and future 6G work. The company also signalled possible future use of the Unitary Patent system, which it does not currently use.
Why this matters: IP positioning for 6G is intensifying, and Indian players are increasingly active. Jio is not only filing patents but contributing directly in 3GPP, including at RAN #111, where it proposed geographical-zone sensing and joined coalitions on agentic AI and NTN requirements. It also tabled a 6G voice baseline paper. Patent strength and standards influence tend to reinforce one another. India’s TSDSI is a 3GPP Organisational Partner, which adds institutional weight to this activity.
Implications for 6G standards: Patents indicate where companies expect to invest and what they expect to be standardised. A growing Jio portfolio strengthens India’s negotiating position as 6G normative work advances under Release 21.
View source material to read more
This survey and tutorial reviews reconfigurable antennas for next-generation mobile networks. It covers fluid, movable, pinching and reconfigurable holographic antennas, which can dynamically alter their radiation patterns and positions. The authors examine channel modelling, performance and resource allocation for each type, and compare their trade-offs.
Why this matters: Antenna flexibility supports the 6G goals of higher gain, better coverage and energy efficiency. Reconfigurable antennas and surfaces are active research inputs to 3GPP, and the work overlaps the reconfigurable intelligent surfaces (RIS) study area. The comparative analysis offers useful guidance for a still-unsettled design space.
View source material to read more
The paper proposes a framework that treats data as tokens for task-oriented image transmission. It combines visual tokens with task tokens generated by a foundation model, coordinated through a cross-modal attention mechanism. The receiver adds an explainable output that highlights which image regions drove a given decision.
Why this matters: Task-oriented and semantic communication is a candidate paradigm for AI-native 6G, moving beyond bit-accurate transmission toward sending only what a task needs. Explainability addresses a real barrier to trusting AI-driven decisions inside networks. The approach links foundation models directly to the air interface, a theme gaining traction in standards discussions.
View source material to read more
This article proposes a hierarchical, cross-domain architecture for converging computing and networking in 6G. It embeds computing resources into the mobile network rather than treating cloud and connectivity as separate layers. Orchestration is handled through hierarchical multi-agent reinforcement learning.
Why this matters: Many 6G scenarios, from immersive communication to integrated AI, demand compute close to the network. Traditional network slicing alone cannot meet these heterogeneous demands. The reported gains in energy use and task satisfaction support the case for compute-network convergence, which recurs in 3GPP architecture study items on computing.
View source material to read more
The paper proposes a channel charting method using physical channel fingerprints for massive MIMO-OFDM systems. It derives compact channel representations from multi-location measurements to build a location-indexed chart. This is then used to acquire statistical channel state information as prior knowledge for channel estimation.
Why this matters: Accurate channel acquisition is a persistent challenge for massive MIMO, which underpins the evolutionary 6G baseline. Location-aware tools like channel charting also support the positioning capabilities prioritised in early 6G. The method reportedly matches traditional online probing while reducing computational complexity.
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.