Every mobile generation starts with a document most people never read. For 5G it was IMT-2020. For 6G it is Recommendation ITU-R M.2160, the IMT-2030 framework, approved in late 2023. It doesn't specify a radio or a protocol. It describes what the system should enable, and it's the clearest picture we have of what 6G is for.
From three triangles to a hexagon
5G was drawn as a triangle: enhanced mobile broadband, ultra-reliable low-latency links and massive machine-type communications. IMT-2030 keeps those three, stretches each one further, and adds three new scenarios. The new ones are what make 6G more than "5G, but faster".
Light circles extend 5G scenarios; filled circles are new in IMT-2030.
The six usage scenarios
- Extends eMBBImmersive communicationHolographic calls, multi-sensory XR, cloud gaming and remote presence that feels like being there. Needs huge, steady data rates with low jitter.
- Extends URLLCHRLLCHyper-reliable, low-latency links for factory automation, remote surgery and cooperating robots, where a late packet is a lost packet.
- Extends mMTCMassive communicationDense sensor fields for smart cities, agriculture and e-health, from hundreds of thousands up to around a hundred million devices per km².
- NewUbiquitous connectivityClosing coverage gaps in rural areas, at sea and in the air. This is where terrestrial towers meet drones, HAPS and LEO satellites.
- NewAI and communicationThe network carries AI (distributed training, inference offloading) and also runs on AI. Think AI-native RAN, digital twins and compute at the cell site.
- NewIntegrated sensing & communicationThe radio becomes a sensor: detecting drones, mapping rooms, recognising gestures, all with the same signals that carry your data.
Four principles behind all of them
The framework also names four overarching aspects that apply to every scenario. I like these because they read like a design brief rather than a feature list:
- Sustainability: energy efficiency treated as a first-class target, not an afterthought.
- Connecting the unconnected: affordable access for the billions still offline.
- Ubiquitous intelligence: AI available everywhere in the system, from devices to core.
- Security, privacy and resilience: networks that protect data and keep working when parts of them fail.
The numbers, next to 5G
M.2160 lists fifteen capabilities: nine enhanced from IMT-2020 and six new ones (coverage, positioning, sensing, AI-related capabilities, sustainability and interoperability). Here are the headline values:
| Capability | IMT-2030 (range) | IMT-2020 |
|---|---|---|
| Peak data rate | 50 – 200 Gbit/s | 20 Gbit/s |
| User-experienced data rate | 300 – 500 Mbit/s | 100 Mbit/s |
| Spectrum efficiency | 1.5 – 3× IMT-2020 | baseline |
| Area traffic capacity | 30 – 50 Mbit/s/m² | 10 Mbit/s/m² |
| Connection density | 10⁶ – 10⁸ /km² | 10⁶ /km² |
| Mobility | 500 – 1000 km/h | 500 km/h |
| User-plane latency | 0.1 – 1 ms | 1 ms |
| Reliability | 1−10⁻⁵ to 1−10⁻⁷ | 1−10⁻⁵ |
| Positioning accuracy | 1 – 10 cm | not specified |
M.2160 gives these as example values to guide research. The binding minimum requirements are being set separately by ITU-R Working Party 5D. Always check the official text before citing.
When does it arrive?
Roughly: the framework was set in 2023. ITU-R is defining requirements and evaluation criteria through 2026. Candidate technologies are submitted and evaluated towards the end of the decade, with specifications around 2030. In parallel, 3GPP started its 6G studies in Release 20 and is expected to produce the first normative 6G specifications in Release 21. So the research being done right now shapes what ends up in the standard.
Where my research fits
BUC-6G, "Beyond Ubiquitous Connectivity", sits where three of the new scenarios overlap: ubiquitous connectivity, AI and communication, and the sustainability and resilience principles that run through them all.
- One fabric from tower to orbit. Towers, drones, HAPS and LEO satellites should behave as one network that steers each service along the best path, not as separate systems glued together.
- AI-native control. O-RAN RICs and AI-RAN compute let learning-based controllers allocate radio, compute and energy in closed loops.
- Zero-touch operations. Intent-based, self-healing orchestration (ETSI ZSM) keeps a network this complex running without an army of operators.
- Energy as a constraint. Every placement and steering decision is also an energy decision.
That's also why the front page of this site goes from a tower up to a satellite. It isn't decoration: it's the system I'm trying to make work as one.