What would move the dates, and how you would know

The timeline has four hinges, and three of them fall within thirteen months

Every date on the list belongs to a body that publishes what happened on it, which makes the watching cheap. Not every date matters equally, though. Four carry most of the risk, because most of what comes later depends on them (Exhibit 1).

The first hinge comes soonest, and it is the one we would least like to miss: the pair of 3GPP plenaries in Boston on 7 to 11 December 2026. RAN#114 has to decide whether any migration path from 5G to 6G beyond spectrum sharing is allowed at all, a question RAN#112 and RAN#113 both left open. In the same week SA#114 receives the architecture decision on the 6G core that the SA plenary has told its working group to reach by November. As if that were not enough for one month, the NTIA’s report on 7.125 to 7.4 GHz, due twelve months after the memorandum of 19 December 2025, lands in December too.

The second is March 2027, when Release 20 is projected to freeze and the Release 21 package is approved, so a slip there is inherited by 6G before 6G has begun. The third is WRC-27 in Shanghai, from 18 October to 12 November 2027, which decides agenda item 1.7 and with it the spectrum. The fourth is the Release 21 Stage 3 freeze in December 2028 at TSGs#122. Counted from 7 October 2026, the four fall at two, five, twelve and twenty-six months. Most of the risk is concentrated in the first year.

The twenty dates in order

The full list runs from an ITU-R Study Group 5 meeting at the end of November 2026 to the day the FCC’s auction authority ends, 30 September 2034. For each row we have named the body that owns it, the document it produces and the part of the series whose table would change (Exhibit 2).

Read down the table and the twenty sort themselves by who owns them. Ten are meetings of 3GPP, the ITU-R or the conference, so the date is the meeting and the outcome is its report or its Final Acts. Seven are deadlines set by statute, by the December 2025 memorandum or by the FCC’s order, and no plenary can move those. That leaves three of a softer kind: the ITU-R’s mid-2030 target, the carriers’ annual guidance and the vendors’ expectation of first commercial systems. The last of those has no body behind it at all.

The timeline moves in three ways, and each way has its own document

Whatever happens on the list, it moves the dates in one of only three ways. The standard can slip, a spectrum decision can go one way or the other, or the carriers can change what they are able to spend. Each has a precedent, a document and a direction (Exhibit 3).

A standards slip is the likeliest of the three and the easiest to see coming. Part 1 puts the base rate at three to nine months, in one direction only, announced at a plenary with a news item. The event we would watch first is the March 2027 freeze of Release 20, because the 6G study has to close before the normative work can be scoped. A slip there is a slip in Release 21 before anyone has touched Release 21. The test could hardly be easier. The 3GPP release page shows a projected date today; on the day, it should show an actual one.

A spectrum decision is the one move that can go either way, and it arrives in two documents about a year apart. The first is the NTIA’s report on 7.125 to 7.4 GHz, due in December 2026 under the December 2025 memorandum, which either identifies the band for reallocation or explains why it cannot. The second is WRC-27, whose Final Acts say whether 7.125 to 7.4 GHz, or more, or nothing, is identified, and for which regions. An identification is less than it sounds. It does not put American spectrum on sale. It allows the NTIA and the FCC to sell it, under a statute that excludes 7.4 to 8.4 GHz and sets July 2033 as the deadline for the federal spectrum.

Either outcome lands in one row of part 6. If the conference does what the carriers asked for in February 2026 and identifies 7.125 to 7.45 GHz, the 7 GHz row stays at 2030 onward. If it does nothing, the question of an American 6G band goes back to the C-band and the upper C-band, which are 5G bands served by 5G radios.

Carrier spending is the slowest signal and the one with the longest record

The three national carriers file their capital expenditure every year under a tag a machine can read. So the shape of the last cycle is on the record, with no estimate of ours in it (Exhibit 4).

The last cycle had a simple shape. The carriers bought spectrum in 2021, their capex peaked in 2022, and it was back near the base within three years. In money, the three spent $44.0 billion in 2019, $56.7 billion at the 2022 peak, the year after the C-band auction, and $47.8 billion in 2025. The equivalent signal for the next cycle is the FY2027 capex guidance, published in early 2028 once the upper C-band has been bought. If it rises the way 2022 did, the budget for 2028 to 2030 is committed to mid-band capacity on existing radios, right through the Release 21 freeze. The 6G build then has to compete for whatever comes after.

Two numbers from part 4 give a sense of how much room there is. After capex and dividends, the three carriers had $33.8 billion left between them in FY2025. At the two recent mid-band prices, the 160 MHz on offer in Auction 115 would cost $36.0 billion to $46.4 billion, which is more than the whole of that surplus. The auction closes in 2027, and the FY2027 releases of early 2028 are the first filings that will show what the carriers mean to spend around it.

Each outcome changes one table in one part

Watching the list costs almost nothing, because the bodies that own the dates publish what they did. Someone has to read the document on the day, and when they have, the change is a single edit to a single table (Exhibit 5).

Six kinds of document carry all twenty outcomes between them, and none of them needs an analyst to interpret it.

1. A 3GPP release page. It shows the word projected beside a freeze date until the freeze happens, then the actual date. Compare the two on the day.
2. A WP 5D meeting report. It lists what was received, so it shows whether the IMT-2030 window opened in February 2027 and closed in February 2029 with 3GPP’s submission inside it.
3. A Final Acts footnote. It names the regions and countries that took an identification at WRC-27, and so settles the ranges in part 2.
4. An FCC public notice. It closes an auction with a gross total and a list of winners, and so settles the price in part 4.
5. An NTIA report. It names a band for reallocation or it does not, and either way it settles the federal position on 7.125 to 7.4 GHz.
6. A Form 10-K. It states the coming year’s capex guidance in a sentence, and so shows whether the 2022 pattern repeats.

The edits are just as contained. A migration decision in December 2026 changes part 8’s map and part 4’s cycle. A Release 21 slip changes the dates in parts 1 and 6 and touches nothing else. A WRC-27 outcome changes the ranges in part 2 and the 7 GHz row in part 6, and an auction result changes the price in part 4. The IMT-2030 requirements, once approved, replace the draft figures in part 3. Nothing on the list rewrites the series.

The terms, briefly

• Plenary. A numbered decision meeting of one of 3GPP’s technical specification groups: RAN#114 is the 114th meeting of the radio group, SA#114 the 114th of the architecture group. The plenary approves work items, freezes releases and publishes a report.
• Projected. The word 3GPP’s release page uses for a freeze date it plans and has not reached. The page replaces it with the actual date when the freeze happens.
• WP 5D. Working Party 5D, the ITU-R group that writes the IMT-2030 requirements and receives candidate technologies. Its meetings are numbered; the submission window runs from #54 in February 2027 to #59 in February 2029.
• Final Acts. The treaty text a World Radiocommunication Conference adopts. Its footnotes to the Radio Regulations name the regions and countries that take an identification.
• Identification. What a conference gives a band when it marks it for IMT in the Radio Regulations. It is permission for administrations to use the band for mobile under agreed conditions; it is not an allocation to any carrier and not a sale.