The petition runs ten pages and makes one request: take 290 megahertz out of the 482 that the Federal Communications Commission intends to open for satellite-to-phone service. It was signed by CTIA, the trade group that speaks for AT&T, Verizon and T-Mobile, and filed days before the commission’s October 29 meeting.
The FCC votes Thursday on a rewrite of its Supplemental Coverage from Space rules. The proposal would add 482 megahertz for satellites that beam signals to ordinary handsets and raise the number of usable bands from five to 13. A second item on the same agenda would auction 25 megahertz in the 1675-1695 and 2020-2025 megahertz ranges for direct-to-device service.
The carriers’ target is narrower than the headline number and larger than it sounds. The 290 megahertz they want excluded sits in the 2.5 GHz and 3.45 GHz bands, which terrestrial 5G networks use heavily. Remove those two blocks and about 60% of the new pool disappears, along with the widest channels a satellite operator could use.
Their argument is not that satellites should stay out of the airwaves. It is that these particular airwaves are built on a timing discipline that a satellite in low orbit cannot share. Both bands run time-division duplex, in which a base station and a handset take turns transmitting on the same frequency. A satellite passing overhead would issue its own timing signals to phones in the same area, and the two schedules would collide.
CTIA told the commission that the conflict cannot be fixed by lowering power limits. The agency’s draft sets an aggregate power flux density ceiling of -76 dBm per square meter per megahertz, a cap on how loudly a satellite may transmit into a terrestrial network. A satellite that respects the cap still arrives at the wrong moment in the frame.
The technical body that writes the world’s mobile standards, 3GPP, has a timing-advance protocol for satellites in its Release 17 and 18 non-terrestrial network specifications. It works in spectrum reserved for satellites, where the satellite alone sets the clock. In shared mid-band, a tower and a spacecraft would both be scheduling the same handset.
SpaceX has spent two years assembling the other side of the argument. In September 2025 it agreed to pay $17 billion for roughly 65 megahertz of EchoStar’s AWS-4 and H-block licenses, ending its reliance on frequencies leased from T-Mobile for the first version of its direct-to-cell service. The FCC approved the transfer in May 2026.
On October 6, the commission’s Space Bureau authorized a constellation of up to 15,000 Starlink satellites built for direct smartphone links on that spectrum. Two days later, SpaceX bought up to 14 megahertz of 800 MHz low-band airwaves from Grain Management, which travel farther and penetrate buildings better than mid-band. On October 21, the FCC cleared its purchase of Tristar’s AWS-3 downlink licenses.
Taken together, the holdings look less like a satellite add-on and more like a carrier’s asset base. Gwynne Shotwell, SpaceX’s president, has said Starlink intends to become a “true mobile service” and expects to take customers from all three national networks. Carrier shares fell between 2% and 5% after a SpaceX presentation in August that laid out the ambition.
The carriers have answered on two fronts. In early October they formed a joint venture to standardize direct-to-device connectivity as an open, neutral layer, a structure that would let them resell satellite capacity without handing the customer relationship to a competitor. And they have taken the spectrum fight to the commission, where the rules are still being written.
What the 482 megahertz would actually change is capacity. Satellite-to-phone service today runs on slivers of licensed airwaves, enough for texts and short voice calls. Wider mid-band channels carry far more data per beam, which is the difference between an emergency fallback and a service people use for video.
That is also why the carriers are not asking for a ban. A secondary allocation, in which satellites must protect terrestrial licensees, would leave the incumbents as the senior users of the band and give them the stronger hand in any commercial arrangement. A primary allocation, or a carve-out that hands SpaceX clear channels, would not.
The commission’s options on Thursday run from adopting the draft as written to stripping the two mid-band blocks and moving the auction forward on its own. The commission has not signaled publicly how it will vote. The draft rules came out on October 8, which gave the carriers three weeks to make their case.
A decision either way will not settle the larger question. Even a favorable vote leaves SpaceX dependent on handsets, towers and backhaul it does not own, and on a standards process where the carriers hold votes. The 3GPP work on shared-band timing is not finished, and no published solution exists for a tower and a satellite scheduling the same phone.
The carriers, for their part, have to decide how much of the orbital business they want to own rather than resist. Their joint venture suggests they have an answer. Whether the commission writes rules that let them act on it is now a matter of five votes in Washington, and a clock that runs on 90-minute orbits.
