Where the spectrum comes from

Congress wrote the pipeline into law, with six dates attached

Public Law 119-21, signed on 4 July 2025 and commonly called the One Big Beautiful Bill Act, restored the FCC’s auction authority in section 40002, and the authority now runs to 30 September 2034. Restoring it was the headline. What matters more for 6G is that Congress, for the first time, attached a quantity of spectrum and a set of dates to it (Exhibit 1).

The requirement comes in two halves, and they add up. The NTIA has to find 500 MHz of federal spectrum between 1.3 and 10.5 GHz and hand it over for reallocation, 200 MHz within two years of enactment and the rest within four. The FCC, for its part, has to auction at least 300 MHz on top of that, and at least 100 MHz of it must come from 3.98 to 4.2 GHz within two years. It must also sell the federal spectrum as the NTIA identifies it, 200 MHz within four years and the remainder within eight. Put the two halves together and you have the 800 MHz.

Two ranges are written out of both halves. Congress kept 3.1 to 3.45 GHz and 7.4 to 8.4 GHz, where federal radar and satellite systems live, out of the covered band and out of the auction authority itself. The 6 GHz band is a different case. Earlier drafts of the bill would have kept it out of the pipeline too, and the law as passed does not.

What is the pipeline worth? The Congressional Budget Office put a figure on it in October 2025: $84.95 billion of reduced direct spending over 2025 to 2034. Most of that, $81.3 billion, is auction proceeds under the restored general authority. A further $3.7 billion comes from the auctions the Act singles out, and the $50 million study appropriation comes off. The Congressional Research Service rounds the whole to over $85 billion. What kind of number is that? A receipts estimate across nine years of auctions, and nothing more; it does not say what any one band will fetch. Part 4 gives the two recent prices that do.

The exclusions deserve as much attention as the targets, and one of them in particular. The 7.4 to 8.4 GHz that Congress kept out is the upper thousand megahertz of the very range WRC-27 is studying for IMT. So the United States wrote itself out of most of the 7 GHz question more than two years before the conference meets. Whatever Shanghai decides about 7.125 to 8.4 GHz, the American part of the answer can lie only in the 275 MHz between 7.125 and 7.4 GHz. The executive branch has since told the NTIA to work on that range, and the carriers and their vendors have asked the FCC to back it. Set the statute’s floors beside the ranges Shanghai is studying and they look small (Exhibit 2).

The FCC moved first, and with 60% more than the law requires

On 22 July 2026 the FCC adopted FCC 26-46, the order that sends 3.98 to 4.14 GHz to auction. That is 160 MHz, which the Commission itself describes as 60% more than the 100 MHz the statute demands. Two days later it gave the sale a number, Auction 115, and a date: bidding starts on 27 April 2027 (Exhibit 3).

The order is concrete in a way the statute is not. It creates 3,248 flexible-use licences across the contiguous United States, eight unpaired 20 MHz blocks in each partial economic area, each with a fifteen-year term. It joins the new band to the 280 MHz sold in 2021, so that what the Commission calls a single harmonised super band runs from 3.70 to 4.14 GHz, 440 MHz in all. A 20 MHz guard band at 4.14 to 4.16 GHz keeps it clear of the satellite downlinks that stay in 4.16 to 4.2 GHz.

The satellite incumbents have to leave, and they leave on two dates. The top 75 licence areas, which held 72.1% of the population at the 2020 census, must be clear by 30 December 2030; the rest by 30 June 2031. Terrestrial service may start the day after each, on 31 December 2030 and 1 July 2031, or sooner in any area whose clearing has been certified. So a winner pays in 2027 for something it can switch on at the end of 2030.

For a 6G plan the useful fact is the band number, and it comes with a footnote. 3.98 to 4.14 GHz lies inside 3GPP band n77, which runs from 3.3 to 4.2 GHz, so the radios built for the 2021 C-band are built for the right band already. The handsets are a different story, and the order treats them as one.

So the upper C-band is new capacity for the generation already in the ground. Whoever wins it can use it from the end of 2030, and nothing about it waits for Release 21. Part 6 places it accordingly.

Everything above 4.2 GHz waits for Shanghai

Every range above 4.2 GHz that 6G is being designed around is on the WRC-27 agenda, and in nobody’s hands. The agenda is narrower, region by region, than the headline ranges suggest, and the American advisory committee has split on it twice (Exhibit 4).

Resolution 256, adopted at WRC-23, is the text behind agenda item 1.7, and the regions in it matter. It invites the 2027 conference to consider identifying 4.4 to 4.8 GHz, but in Regions 1 and 3 only. The 7.125 to 8.4 GHz range is on the table in Regions 2 and 3 in full, and in Region 1 only at its two ends, 7.125 to 7.25 and 7.75 to 8.4 GHz. The third range, 14.8 to 15.35 GHz, is on the table everywhere. The United States is in Region 2. Which is another way of saying that 4.4 to 4.8 GHz is off the American table altogether, and could reach it only by a footnote of its own, while 7.125 to 8.4 GHz is on it in full.

The FCC’s WRC-27 Advisory Committee could not agree a single view on 7 GHz. In February 2026 it sent the Commission three of them, in DA 26-182, with the NTIA’s own draft attached.

• View A. CTIA, AT&T, Ericsson, GSMA, Nokia, Qualcomm, T-Mobile and Verizon want 7.125 to 7.45 GHz identified.
• View B. Apple, Boeing, Broadcom, Charter, Comcast, Lockheed Martin, NCTA, the Satellite Industry Association, the UWB Alliance and four others want no change in 7.4 to 8.4 GHz. They call any position on 7.125 to 7.4 GHz premature while the studies are pending.
• View C. The View A companies, minus CTIA, would go further and ask for 7.125 to 8.4 GHz or portions of it.
• The NTIA’s draft. 7.125 to 7.4 GHz, with the upper figure in brackets.

In August the committee split again, in DA 26-835, this time over 4.4 to 4.8 GHz. The carriers and vendors asked for 4.5 to 4.8 GHz to be identified; three aviation and aerospace members opposed any change at all, on radio-altimeter grounds. The Commission’s own words on every one of these views are that they do not constitute final United States positions. So the industry has not agreed among itself, and the Commission has not committed.

The federal side has a clock of its own. National Security Presidential Memorandum 8, issued on 19 December 2025, directs the NTIA to begin identifying 7.125 to 7.4 GHz for reallocation under section 40002. It also orders a study of where the federal systems in that range could go instead, 7.4 to 8.4 GHz included, with a report due within twelve months. That puts the report in December 2026, and in May the NTIA’s administrator told a conference it was on track. The band is one of four on the NTIA’s published study list. There is an older and wider study too, of all of 7.125 to 8.4 GHz, which the National Spectrum Strategy began in March 2024; it is due to report in October 2026.

So the American 7 GHz question runs in sequence rather than in parallel. First an NTIA report at the end of 2026, then a conference decision in November 2027, then an auction the statute allows until July 2033 (Exhibit 5).

It helps to remember what the last conference did with the band just below. WRC-23 identified 6.425 to 7.125 GHz for IMT across Region 1 and in Brazil and Mexico. It identified 7.025 to 7.125 GHz across all of Region 3, and 6.425 to 7.025 GHz in Cambodia, Laos and the Maldives. Those identifications are footnotes 5.457D to 5.457F of the Radio Regulations, and the United States is named in none of them. It had opened the whole of 5.925 to 7.125 GHz to unlicensed use in April 2020, and it went to Dubai proposing no change. So the first 6G-era band in Europe, if the identification is used, is one the American market has given to Wi-Fi.

Published Radio Regulations, 2024 edition, Article 5 · ITU closing release, 15 December 2023 · GSMA summary, February 2024 · FCC 20-51, 23 April 2020

Every range on the agenda costs more signal than the C-band

There is a cost here that no regulator can waive. Free-space path loss rises with the square of frequency, so every one of the ranges above costs more signal than the band 5G was priced on. Against 3.5 GHz, the middle of that band, 7.125 GHz costs 6.2 dB and 15.35 GHz costs 12.8 dB (Exhibit 6).

The relation is equation 6 of Rec. ITU-R P.525-5: 20 log₁₀ of the frequency ratio, in decibels. It is the floor on the extra loss, before anything the environment adds. Run the ranges through it. The top of the upper C-band, 4.14 GHz, costs 1.5 dB against 3.5 GHz, which is hardly anything. At 4.8 GHz, the top of the first WRC-27 range, it is 2.7 dB, still modest. Then comes the step: 7.125 GHz costs 6.2 dB, 7.45 GHz 6.6 dB, and 15.35 GHz 12.8 dB. Six decibels is a factor of four in power. In free space it halves the distance at which the same link closes.

Walls, it turns out, charge less for the move in frequency than free space does, and a great deal more for something else. Rec. ITU-R P.2109-2 models building entry loss for two classes of building. For traditional construction, its median at horizontal incidence rises from 15.7 dB at 3.5 GHz to 17.0 dB at 7.125 GHz. For thermally efficient construction, the kind with metallised glass and foil-backed panels, it rises from 30.9 dB to 32.7 dB. So the move to 7 GHz costs one to two decibels indoors, and the kind of building you are trying to reach costs about fifteen. WInnForum’s measurements at 6.5 GHz in 2021 found the same two populations, at 17.8 dB and 33.7 dB, which gives us some confidence that the model is not far off.

Before any new feature of the standard earns a cent, a 6G coverage plan has to pay back 6 dB between the C-band and 7.125 GHz. It pays with antenna gain, with more sites, or with lower expectations, and the bill does not change with whatever happens at Shanghai. It is also why the upper C-band, at 1.5 dB, is worth more to a near-term plan than its width alone suggests.

Put the statute, the order, the agenda and the equation together and the spectrum story has a shape. The next 160 MHz is American and mid-band. It is sold from April 2027 and usable from the end of 2030, on radios already built for the band, with handsets the order requires to follow. The 7 GHz range that 6G is designed around is, for now, an NTIA report in December 2026, a conference decision in November 2027 and an auction the statute allows until 2033. It comes with a 6 dB bill attached. Anything a roadmap says about 6G coverage before those dates is, whatever the label on the slide, a statement about the C-band.

The terms, briefly

• Covered band. The statute’s name for 1.3 to 10.5 GHz less the two excluded ranges. The 500 MHz the NTIA identifies has to come from it.
• Partial economic area. One of 416 licence areas the FCC uses for mid-band auctions. The top 75 by population are the first to be cleared and served.
• Region 1, 2, 3. The three ITU regions, roughly Europe and Africa, the Americas, and Asia and the Pacific, in which the Radio Regulations can differ. An identification can apply to one region, to all three, or to named countries by footnote.
• MHz-POP. Megahertz of bandwidth multiplied by the population of the licence area. Auction prices are compared in this unit, and direct-to-device part 4 prices the recent sales in it.