Daily Dose of 6G

The physical layer 6G decided not to reinvent

The RAN1 radio study is a live draft — TR 38.760-1 reached v0.4.0 on 14 August, and it does not close until June 2027. What the reported agreements so far describe is not a clean sheet. It is 5G NR, carefully re-argued.

Transcribed from 3GPP’s published Release 21 timeline (© 3GPP 2026). The highlighted row is where TR 38.760-1’s conclusions become specification — not this year, but Q3 2028.

The document, and the hand that holds it

The 3GPP portal record for TR 38.760-1 is short and worth reading literally. Title: “Study on 6G Radio RAN1 aspects” — the plain wording, not a descriptive gloss. Series 38. Responsible group RAN1. Initially planned release: Release 20. Rapporteur Shinya Kumagai, NTT DOCOMO INC. Latest version 0.4.0, uploaded 14 August 2026. Linked work item FS_6G_Radio, ID 1080072, WID RP-261559, running 6 June 2025 to 4 June 2027.

Two of those facts matter more than the rest. A v0.x draft means nothing in the document is concluded; the TR is a container that fills up over two years and is only argued as a whole at the end. And a June 2027 end date puts the radio study’s close after the March 2027 approval of the Rel-21 package — so the 6G Radio work item will be chartered while its own study is still three months from finishing.

One find the catalog did not have: the portal lists five deliverables against work item 1080072, not four. Alongside TR 38.760-1 through -4 sits TR 38.960, “Study on 6G Radio”, owned by the RAN plenary rather than a working group — the umbrella document above the four per-group parts.

Waveform: continuity, argued rather than assumed

The single most-watched question in 6G radio design was whether OFDM survives. On the reporting available, it does. Ericsson’s standardization blog, writing after the June 2026 Singapore plenary, states the position without hedging: “3GPP has agreed to use cyclic-prefix orthogonal frequency-division multiplexing (CP-OFDM) in the downlink. For uplink, two waveforms are supported: CP-OFDM and discrete Fourier transform spread OFDM (DFT-s-OFDM).”

Qualcomm’s Juan Montojo, writing in RCR Wireless on 1 July 2026, frames the same outcome as continuity and adds the timing: the baseline decisions on waveform, modulation, frame structure and channel coding were reached in RAN1 in August and October 2025, not in Singapore. The plenary consolidated what the working group had already settled a year earlier.

Compiled from Ericsson’s and Qualcomm’s published summaries, June–July 2026. These are vendor accounts of RAN1 agreements, not text quoted from TR 38.760-1, which this newsletter has not read.

What “reuse” is actually buying

It is tempting to read continuity as a failure of ambition. The better reading is that the physical layer is no longer where a generation differentiates itself, and RAN1 appears to have accepted that early rather than spending two years relitigating it.

Montojo’s account gives one concrete number for the new mid-band: around 7 GHz, up to 400 MHz channel bandwidth from the network side, and up to 400 MHz downlink / 200 MHz uplink from the device side. That asymmetry is a device-cost decision dressed as a bandwidth figure, and it is the kind of thing RAN4 will spend Editions 12 and 13 arguing about.

Where this is being decided — including this week

RAN1 meets roughly six times a year, and the agreements above are traceable to specific rooms. The portal’s RAN1 calendar places RAN1#122 in Bengaluru, 25–29 August 2025, and RAN1#122-bis in Prague, 13–17 October 2025 — the two meetings Montojo’s “August and October 2025” points to.

As this edition goes out, RAN1#126 is in session in Maastricht, 24–28 August 2026. Between now and the study’s close there are five more: #126-bis in South Korea (12–16 October 2026), #127 in Calgary (16–20 November 2026), #128 in Osaka (25–29 January 2027), #128-bis in Lyon (19–23 April 2027) and #129 in China (24–28 May 2027) — the last of which sits eight days before the 4 June deadline. Six meetings is the entire remaining budget for a radio design.