Analysis · 6G · part 3 of 8
What 6G adds, and what 5G-Advanced already does
Every 6G claim traces back to one ITU document, and that document gives its numbers as targets for research rather than as requirements. The releases that make up 5G-Advanced, by contrast, have work-item identifiers and devices in the shops. Put the two side by side and four of the six 6G scenarios already have a 5G-Advanced pathway, while the one draft 6G requirement that has reached the public asks for less than the headlines.
Sources: Recommendation ITU-R M.2160-0 (11/2023), read in the published text, with the ITU’s release of 1 December 2023 and RCR Wireless’s summary as second sources; Report ITU-R M.2410-0 (11/2017); the ITU’s note of 17 March 2026 on the draft minimum requirements, the APT’s WP 5D update by the IMT-TECH sub-group chair and Samsung Research’s account; 3GPP’s Release 17, Release 18 and Release 19 pages and its Highlights of 2021 on the 5G-Advanced name; the ATIS and 3GPP webinar deck of 3 April 2024, with each work and study item checked against the RAN#103 document list; the GSA’s RedCap note on 3gpp.org and an overview of URLLC in Releases 15 and 16; T-Mobile’s newsroom on 16 October 2024 and 9 December 2025, AT&T’s 16 July 2025 post, and RCR Wireless and Fierce (17 October 2024, 16 July 2025, 19 September 2025) as second sources. Retrieved and cross-checked 2 Oct 2026.
The framework gives targets for research, and says so
Recommendation ITU-R M.2160, approved in November 2023, is where every 6G claim has to begin, because it is the document that defines what IMT-2030 means. It sets out six usage scenarios, four overarching aspects and fifteen capabilities. For the capabilities it gives numbers, but it is careful about what they are: estimated targets for research and investigation, in its own words (Exhibit 1).
The six scenarios are immersive communication, hyper reliable and low-latency communication, massive communication, ubiquitous connectivity, artificial intelligence and communication, and integrated sensing and communication. The first three are IMT-2020’s three, extended. Two, AI and sensing, are new as scenarios. Ubiquitous connectivity is the satellite and remote-area case, which the direct-to-device series covers. Across all six run four aspects: sustainability, security and resilience, connecting the unconnected, and ubiquitous intelligence.
Of the fifteen capabilities, the framework counts nine as enhanced from IMT-2020. Those are the eight that had numbers in 2017, plus security and resilience. The other six it counts as new: coverage, positioning, sensing, AI, sustainability and interoperability. Only one of the new six, positioning, arrives with a number attached.
| Capability | IMT-2020 (M.2410, 2017) | IMT-2030 (M.2160, 2023) |
|---|---|---|
| Peak data rate | 20 Gbit/s down, 10 up | Greater; 50, 100, 200 Gbit/s given as possible examples |
| User experienced data rate | 100 Mbit/s down, 50 up (dense urban) | 300 and 500 Mbit/s given as possible examples |
| Latency, user plane | 1 ms URLLC; 4 ms eMBB | 0.1 to 1 ms, research target |
| Reliability | 1 − 10−5 | 1 − 10−5 to 1 − 10−7 |
| Connection density | 106 per km2 | 106 to 108 per km2 |
| Area traffic capacity | 10 Mbit/s/m2 (indoor hotspot) | 30 and 50 Mbit/s/m2 as possible examples |
| Mobility | 500 km/h | 500 to 1,000 km/h |
| Peak spectral efficiency | 30 bit/s/Hz down, 15 up | 1.5× and 3× IMT-2020 as possible examples |
| Positioning accuracy | Not a minimum requirement | 1 to 10 cm, research target |
| Coverage, sensing, AI, sustainability, security and resilience, interoperability | Not capabilities | Described without numeric targets; sustainability names its metric, bit per joule |
IMT-2020’s values are minimum requirements a candidate had to meet. IMT-2030’s are the framework’s research targets, which the Recommendation says may not be reached simultaneously in any one scenario. The two columns are not the same kind of number.
Source: Report ITU-R M.2410-0, November 2017; Rec. ITU-R M.2160-0, November 2023, section 4; ITU release, 1 December 2023 (Published).
Set the two columns side by side and the multiples are less dramatic than the headlines. The example peak rate is two and a half to ten times IMT-2020’s, and the example user rate three to five times. Latency improves tenfold at the bottom of its range and not at all at the top. Positioning is the only capability that is wholly new as a number, at one to ten centimetres. And the two capabilities that are wholly new as categories, sensing and AI, have no number at all. That is where the standard still has the most to decide, and the Release 20 study is the place it is deciding it.
The draft minimum requirements ask for less than the targets
In February 2026 Working Party 5D turned the framework’s targets into a first draft of minimum requirements. Only two of its figures have reached the public so far, and both of them are close to what 5G-Advanced does already (Exhibit 2).
The draft lists twenty technical performance requirements. How many of them are new depends on who is counting: the ITU says seven, and the delegations that have published the list say six, naming composite requirements, link distance, positioning, sensing, AI, and resilience and extended connectivity. The full table stays a working document until Study Group 5 approves it on 1 December 2026. The APT’s meeting update gives the two figures we have. The theoretical peak rate is 36 Gbit/s down and 18 Gbit/s up, and that assumes 600 MHz of bandwidth. User-plane latency is 4 ms for immersive communication and 1 ms for hyper reliable and low-latency communication.
Source: Report ITU-R M.2410-0 (11/2017); APT, WP 5D updates as of WP 5D#51, March 2026; Rec. ITU-R M.2160-0 (11/2023) (Published). Note: the draft minimum is a theoretical peak assuming 600 MHz of bandwidth and remains a working document until Study Group 5 approves it.
Hold those two figures against the framework and the distance between the generations shrinks. A peak of 36 Gbit/s in 600 MHz is 1.8 times IMT-2020’s 20 Gbit/s, where the example values had suggested anything from two and a half to ten times. The 1 ms latency requirement for the reliable scenario is the figure 5G has to meet today. So the minimum is being set near the top of what 5G-Advanced does, and the headline multiples live in the research targets above it. A product that meets the minimum will be 6G. A product that meets the targets is the research programme.
Four of the six scenarios already have a 5G-Advanced work item
3GPP’s Project Coordination Group approved the name 5G-Advanced in April 2021, for Release 18 onward. Release 19, the second 5G-Advanced release, froze in December 2025, and it is the last whose content was settled before the 6G study opened. Read its work items with the IMT-2030 scenarios in mind and they look like a first draft of several of them (Exhibit 3).
Release 18’s page lists three new topics beside the continuing ones: energy efficiency, artificial intelligence and machine learning, and extended reality. Release 19’s radio projects were laid out in the ATIS and 3GPP webinar of April 2024, each with an identifier that resolves to a document on 3GPP’s server. Its Stage 3 froze in September 2025 at SA#109 and its code on 12 December 2025 at SA#110, which are the dates the webinar had published twenty months earlier. 3GPP, in other words, did what it said it would do, when it said it would.
- AI and machine learning. A normative work item for the NR air interface (RP-240774), which was the largest project in the release, and a study on AI for mobility (RP-240082).
- Sensing and new bands. Channel-model studies for integrated sensing and communication (RP-240799) and for 7 to 24 GHz (RP-234018).
- Devices and power. A study on ambient IoT (RP-240826), work on low-power wake-up signals and receivers (RP-240801), and network energy saving enhancements (RP-240170).
- Radio capacity. A fifth phase of MIMO (RP-240087) and subband full duplex (RP-240789).
- Services. A third phase of XR (RP-240791), and a third phase of non-terrestrial networks with regenerative payloads and RedCap over satellite (RP-240775).
| IMT-2030 scenario | Release 17 to 20 work on it | State on 2 Oct 2026 |
|---|---|---|
| Immersive communication | XR phases 1 to 3 (Rel-18; Rel-19 RP-240791); MIMO phase 5 (RP-240087) | Specified; devices and services are the carriers’ to report |
| Hyper reliable, low latency | URLLC from Rel-15, enhanced in Rel-16; Rel-19 duplex evolution (RP-240789) | Specified to 1 ms, the draft 6G minimum; sub-millisecond is a research target |
| Massive communication | RedCap (Rel-17), eRedCap (Rel-18), ambient IoT study (Rel-19 RP-240826) | Commercial: RedCap on T-Mobile since 17 October 2024; AT&T nationwide since 16 July 2025 |
| Ubiquitous connectivity | NTN from Rel-17; Rel-19 phase 3 (RP-240775) with regenerative payloads | Commercial, by other routes: see the direct-to-device series |
| AI and communication | Rel-18 new topic; Rel-19 air-interface work item RP-240774 and mobility study RP-240082; Rel-20 two-sided model work item RP-251870 | Specified in Rel-19 for one-sided models; no operator has published a deployment |
| Integrated sensing and communication | Rel-19 channel model study RP-240799; Rel-20 work item RP-261566 (June 2026), base-station drone detection only | One narrow work item; the sensing capability M.2160 describes waits for Release 21 |
Commercial means an operator has published that the feature is in service. The absence of such a publication is not evidence that a feature is unused, only that it has not been announced.
Source: 3GPP Release 17, 18 and 19 pages; ATIS and 3GPP webinar deck, 3 April 2024, checked against the RAN#103 document list; 3GPP RP-261566 (RAN#112, June 2026) and RP-251870; T-Mobile newsroom, 16 October 2024; AT&T, 16 July 2025 (Published).
RedCap is the clearest case of a 5G-Advanced feature with a date and a device behind it. It is a Release 17 device class, frozen in March 2022, and it put the massive-communication scenario into service more than four years before the first 6G specification will exist.
Both generations have dated milestones, and the two sets do not overlap (Exhibit 4). Every 5G-Advanced milestone, from the naming in 2021 to the Release 19 code freeze, has a month against it. The first 6G document with the force of a requirement is due for approval on 1 December 2026, and the first 6G specifications freeze two years after that.
Source: 3GPP release pages and Highlights 2021; ATIS and 3GPP webinar deck, 3 April 2024; Rec. ITU-R M.2160-0; ITU note, 17 March 2026; T-Mobile, 16 October 2024; AT&T, 16 July 2025; 3GPP RP-261566 (Published). Note: the March 2027 and December 2028 dates are 3GPP’s current plan, as part 1 of this series records.
The line between the generations runs through the middle of most capabilities
Read from these documents, the line between the generations does not fall between capabilities. It runs through the middle of most of them. Immersive, massive and ubiquitous communication are 5G-Advanced features with 6G targets at the top of their range. AI in the air interface is a Release 19 specification whose 6G form is being studied. Sensing has a channel-model study, one Release 20 work item for drone detection by base stations, and a 6G form that waits for Release 21. The only capabilities that live wholly on the 6G side of the line are the ones with the smallest numbers attached: a tenth of a millisecond, a centimetre.
Four of the six IMT-2030 scenarios have a Release 17, 18 or 19 work item that addresses them by name, and two of those four have a commercial service behind them. Sensing has a channel-model study and one Release 20 work item, limited to drone detection by base stations. The sub-millisecond, centimetre-grade end of latency and positioning has nothing before Release 21. Those two are what the word 6G buys that 5G-Advanced does not, and the draft minimum requirements, where public, fall on the 5G-Advanced side of that line.
For a business case this matters in a particular way. A plan that lists immersive, massive or ubiquitous communication as 6G revenue is listing revenue that Release 19 equipment can earn now, and it ought to say why it is not earning it. A plan that lists sensing or sub-millisecond control is listing revenue that waits for the Release 21 freeze and a radio designed to it, which puts its first date in the next decade. Part 6 sorts the candidates on that basis, and the integrated sensing series sets out what the sensing standard has agreed so far.
The terms, briefly
- Usage scenario. The ITU-R’s name for a family of uses a generation is designed around. IMT-2020 had three; IMT-2030 has six. A scenario is a heading, and a heading is not a feature.
- Research target and minimum requirement. The framework gives targets for research. The later Report gives the minimum a candidate technology has to demonstrate to be recognised as IMT-2030. Only the second is a pass mark.
- Work item and study item. A 3GPP study item produces a technical report and no specification text. A work item produces normative text. RP-numbered documents are the RAN plenary’s approvals of each.
- RedCap. Reduced capability, a Release 17 device class with fewer antennas and narrower bandwidth, for wearables and sensors. eRedCap in Release 18 caps the peak rate at 10 Mbit/s to cut cost further.
Implications
Carrier strategist
Audit the 6G slide against Exhibit 3. Anything in its first four rows is a 5G-Advanced feature with a work item behind it and, in two cases, a competitor’s commercial service. For those rows the 6G question is one of timing on hardware you already own, and a new network does not come into it.
Vendor
Quote the Release 19 work item and say what the product does with it, because that is a claim a buyer can check. The framework’s numbers are research targets, and the one published draft minimum is 36 Gbit/s in 600 MHz. A datasheet that quotes 200 Gbit/s as a 6G requirement is quoting a target.
Investor
Attribute no revenue before 2030 to sensing or to sub-millisecond control. Those two scenarios have no commercial record and no spectrum, while the four that are partly 5G-Advanced have both. Treat a 6G revenue line dated earlier than that as a claim about 5G-Advanced.
Method and limits
How this was built
The IMT-2030 capability values are the ranges and example values the framework Recommendation gives, read from its published text. Where the framework calls a value a possible example or a research target, we keep that word, because the distinction is the point. The IMT-2020 values are the minimum requirements in Report M.2410. We quote the draft IMT-2030 minimum requirements only where a public document gives a figure, which at the time of writing means the APT’s meeting update. The 5G-Advanced work items are named with their 3GPP identifiers from the ATIS webinar deck, and we checked each identifier against the RAN#103 document list.
We count a scenario as having a 5G-Advanced pathway when a Release 17, 18 or 19 work item or study item addresses it by name. We show commercial status only where an operator has published it. The Release 20 sensing and AI work items are taken from part 8 of this series, which reads the plenary documents.
What it does not show
The IMT-2030 minimum requirements completed in draft in February 2026 are an ITU-R working document until Study Group 5 approves them. The full table is not public, and we quote only the figures the APT and Samsung have published. A work item in a frozen release means a specification exists; it does not mean any network has turned it on. Verizon’s RedCap date has no newsroom release behind it, so we do not give one.
Data as of: ITU-R and 3GPP documents as read 2 Oct 2026 · Method version 1.1 (fact-checked against second sources).
Found an error? Tell us. Corrections are published on the piece that carried them.