The FCC just licensed a cell network in orbit: what SpaceX's 15,000-satellite direct-to-phone approval really means
On Oct. 6 the FCC cleared SpaceX to launch 15,000 satellites that talk straight to ordinary phones, using airwaves SpaceX is buying from EchoStar and without needing a carrier partner. Here's what the order really says, what it could mean for people in dead zones, carriers and the Pentagon, why it all depends on Starship, and what another 15,000 satellites means for an increasingly crowded low Earth orbit.
Picture this. It’s late afternoon on a ranch road in the high country, an hour from the nearest cell tower. A calf is down in a ravine, a pickup has a flat, and the phone in her jacket shows the two words every rural driver knows by heart: No Service.
Then the phone finds a signal anyway. It isn’t coming from a tower. It’s coming from a satellite racing across the sky at nearly five miles a second, a couple hundred miles up. Today that link is good for a text message, maybe a little data. In the world the FCC just signed off on, it could carry a clear voice call or a live video feed to the vet, on the same phone she bought at the mall.
That rancher is made up, but the dead zone isn’t. Anyone who’s driven the back roads of rural America knows one. On Tuesday, Oct. 6, the Federal Communications Commission took a big step toward closing them. In Authorization and Order DA 26-1078, the FCC gave SpaceX permission to launch and run 15,000 new satellites built to talk straight to ordinary phones. They’ll use radio spectrum SpaceX is buying from EchoStar, and for the first time SpaceX won’t need a wireless carrier’s blessing to use it.
Most headlines stopped at the big number. The more interesting story is in the fine print: who controls the airwaves, what the FCC waved through, what it set aside for later, and how fast SpaceX has to build. Then there’s a question a lot of readers have been asking: how crowded is low Earth orbit getting, and what happens when you add 15,000 more satellites?
So what exactly did the FCC say yes to?
The order comes from two FCC offices, the Space Bureau and the Wireless Telecommunications Bureau, under GN Docket 25-340. Officially it “grants in part” and “defers in part.” SpaceX got most of what it asked for, with strings attached and a few requests parked for later.
Here’s the shape of it. The 15,000 satellites will fly in nine “shells,” which are rings of orbits at slightly different heights and tilts. Those shells run from 326 to 335 kilometers up (about 203 to 208 miles), at angles from 26 degrees all the way to a near-polar 96.87 degrees (FCC order; Via Satellite). That’s very low, well below most of the Starlink internet satellites flying today.
Inside the U.S., the satellites can offer two kinds of service. The first is “Supplemental Coverage from Space,” the FCC’s term for satellites borrowing cellular airwaves to fill in where towers can’t reach. The second is traditional mobile satellite service. Outside the U.S., they can offer mobile satellite service and what the industry calls “direct-to-cell.” They’re also cleared for high-frequency links back to ground stations for carrying traffic and controlling the fleet.
The airwaves are the heart of it. The satellites can use four slices of spectrum in the U.S.:
- 1695–1710 MHz, for phones talking up to the satellite
- 1915–1920 and 1995–2000 MHz, known as the AWS-H Block
- 2000–2020 and 2180–2200 MHz, the 2 GHz band, mostly for satellites talking down to phones
- T-Mobile’s PCS G Block at 1910–1915 and 1990–1995 MHz, which already powers today’s service
All told, that’s 15 MHz of new upload capacity “along with the 50 megahertz of D2D downlink” plus the T-Mobile lease. In the FCC’s words, that “ensures SpaceX can provide meaningful two-way D2D service.” (D2D just means direct-to-device.)
Then there’s the clock. SpaceX has to get half the fleet, 7,500 satellites, in orbit and working by Oct. 7, 2032, and all 15,000 by Oct. 7, 2035. The license runs 15 years, starting the day the first satellite reaches its assigned orbit. SpaceX owes the FCC a surety bond, essentially a financial guarantee it will follow through, by Nov. 4, 2026. It also has 30 days to turn the conditions down if it doesn’t like them (Light Reading). Twice a year, by Jan. 1 and July 1, it has to report close calls with other objects, the dodges it made, satellites it pulled from service, and how many came back down through the atmosphere.
FCC staff didn’t hide their enthusiasm. The order says the network “will reshape spectrum-based connectivity by improving competition, allowing more consumers more choices in more places.” It predicts the satellites will help “initially by supplementing service availability in hard-to-reach geographic areas of the country and ultimately through increased competition in the provision of retail wireless services.”
The one waiver that changes everything
Under the FCC’s 2024 rules, a satellite company that wants to use cellular airwaves has to lease them from a carrier like T-Mobile or AT&T. That’s the deal SpaceX has with T-Mobile today.
This time, though, SpaceX will own the ground licenses for the AWS-3 and AWS-H airwaves itself. Making it sign a lease with itself would be silly, and the FCC said so: “Requiring SpaceX to enter into a leasing arrangement with itself would be an unnecessary regulatory burden.” PCMag and TechSpot both point out what that means in practice. SpaceX now has a regulatory path to sell phone service on its own, in competition with AT&T, Verizon and even its current partner, T-Mobile.
There are strings. The waiver can’t be passed along if SpaceX ever sells the spectrum. And SpaceX has to follow the FCC’s interim 911 rules “as if SpaceX were a CMRS provider,” meaning as if it were a regular cell carrier. If you call or text 911 through a satellite, it still has to reach the right local dispatcher.
A few other waivers make the design work. The FCC let SpaceX flip the 2000–2020 MHz band to mostly downlink across most of the country, agreeing that “mobile broadband uses far more downlink than uplink spectrum.” In plain terms, we all watch and download far more than we upload. It let satellites, instead of ground towers, tell phones when they’re allowed to transmit on the AWS-3 airwaves. And it allowed a bit more stray signal at the edges of SpaceX’s AWS-H and PCS G channels: −110.6 dBW/m²/MHz instead of the usual −120, within five megahertz of the band edge.
What SpaceX didn’t get, at least not yet
The FCC held back on several things:
- It deferred SpaceX’s request to use 2020–2025 MHz for uplink here at home.
- It deferred SpaceX’s much broader request for direct-to-cell service abroad across the whole 1429–2690 MHz range.
- It deferred a pair of high-frequency bands abroad (20.2–21.2 and 30.0–31.0 GHz) “in light of the extensive federal use.”
- Requests for the Big LEO bands (1610–1626.5 and 2483.5–2500 MHz) and L-band frequencies outside the U.S. had already been thrown out in an earlier order. That made the Globalstar and Iridium protests against those bands moot.
- Viasat’s petition to block the application was denied in part.
The biggest condition is about timing. SpaceX can only test on the EchoStar airwaves until the second and final step of the EchoStar deal closes, expected around Nov. 30, 2027 (PCMag; Via Satellite).
There’s a twist hiding in that 2020–2025 MHz deferral. On Oct. 29, Chairman Brendan Carr’s FCC is set to vote on auctioning 25 MHz in the 1675–1695 MHz and 2020–2025 MHz bands for satellite-to-phone service, ground networks, or a mix of both. The same meeting takes up new rules that would open another 482 MHz of airwaves for carriers to lease to satellite companies (Light Reading). Our read: the slice the FCC just withheld from SpaceX is one of the slices it plans to auction. If SpaceX wants it, it may have to win it.
How do you get a phone to talk to a satellite?
Start with the airwaves. SpaceX is buying EchoStar’s AWS-4, AWS-H Block and unpaired AWS-3 licenses, about 65 MHz in all. The FCC approved that purchase in May (DA 26-471), on the condition of a $2.4 billion escrow tied to EchoStar’s abandoned plans to build a 5G network (SpaceNews). The deal started at about $17 billion and grew as SpaceX added more EchoStar spectrum. Citing SpaceX’s IPO filings, Broadband Breakfast puts the total at $19.6 billion, with closing expected next year.
Then there’s the hardware. Your phone’s antenna is tiny and weak, so the satellite has to do the heavy lifting. At Mobile World Congress in March, SpaceX’s Michael Nicolls said each second-generation “V2” satellite has an antenna five times larger than the first generation’s, 16 times as many beams, and “nearly 100 times the data density” (PCMag). SpaceX is aiming for speeds of up to 150 Mbps per user. Today’s network of more than 650 direct-to-cell satellites delivers about 4 Mbps (Light Reading).
Each satellite acts like a cell tower in the sky, following the global cellular standards body’s (3GPP) rules for “non-terrestrial networks,” so phones don’t need special parts. The order says each beam will be “roughly 12 km” wide, about the size of a small town. That tight focus lets SpaceX switch phone transmissions on or off neighborhood by neighborhood, which is how it plans to protect sensitive government sites and avoid stepping on other license holders.
The low altitude matters too. Our analysis, not the FCC’s: at about 330 kilometers, the satellite sits a lot closer to your phone than satellites flying at typical internet-constellation heights. That gives a stronger signal and tighter beams. The catch is that there’s still a whisper of atmosphere up there, so the satellites have to keep firing their thrusters and get replaced on a steady cycle. The upside of that drag is that a dead satellite doesn’t hang around for long. More on that below.
Why should anyone outside the telecom world care?
Satellite phones stop being just for emergencies. Today’s service began with emergency alerts and simple texts. In October 2024, after Hurricanes Helene and Milton, the FCC gave SpaceX and T-Mobile emergency permission to beam alerts to phones and enable texting, including texts to 911, in hard-hit areas (Reuters). By March, SpaceX said about 10 million people a month were using the service and expected more than 25 million by the end of 2026 (Teslarati). Nicolls says V2 should feel like a 5G network, with calls, video and data on unmodified phones, and that at launch it’ll be “on most devices in the US” (PCMag).
Think about who that helps. A sailor a few miles off a rocky coast. A wildland fire crew in a canyon where radios bounce off the walls. A family whose town just lost every tower in a flood. (These are illustrations, not specific cases we’re reporting.) For them, the gap between “a text might get through” and “I can video-call for help” isn’t a spec sheet. It’s whether help can see what they see.
It’s going global. Carriers abroad are already signing up for V2. Deutsche Telekom plans service in 10 European countries starting in 2028 (TelecomTV), and Japan’s KDDI is aiming for the end of 2027 (TechSpot). TelecomTV notes V2 is built around “globally harmonised S-band spectrum,” meaning airwaves set aside for similar use in many countries. That makes it easier for the same phones to work almost anywhere.
It shifts the balance with the carriers. With its own airwaves and no lease requirement, SpaceX doesn’t have to depend on any one carrier. Gwynne Shotwell, SpaceX’s president, has called next-gen Starlink Mobile “100 times better than what we’ve got right now” and says service should start by the end of 2027 (Via Satellite). CFO Bret Johnson said last month SpaceX could “stand up terrestrial solutions” or work with existing carriers (Broadband Breakfast).
Where does the Pentagon fit in?
Let’s be clear up front: this is a commercial license. Nothing in it gives SpaceX a military job. But if you work in defense, three things should catch your eye.
First, the FCC built protections for government users into the license. The 1695–1710 MHz upload band is shared with 47 federal ground stations, mostly sites that receive weather-satellite data, guarded by 27 protection zones. SpaceX can’t let phones transmit inside any of those zones until it signs a written coordination agreement with the federal users. Before it uses a long list of higher-frequency bands near federal ones, it has to trade technical details with the agencies involved “consistent with national security requirements.” And it has to keep a 24/7 contact on call for NTIA, the agency that manages federal spectrum. Our read: this shows how much of America’s best mid-band spectrum is shared with government, and how much the FCC is counting on those 12 km beams to keep everyone out of each other’s way.
Second, the Pentagon has already wanted this. CNBC and Reuters reported in May that during the Iran conflict, Pentagon officials talked with SpaceX about using direct-to-cell service to help Iranian civilians get around government blackouts and jamming, connecting “without terminals on the ground.” According to Pentagon documents Reuters reviewed, SpaceX proposed charging up to $500 million to set it up and $100 million a month to run it (CNBC/Reuters). That was before V2 existed. It’s a strong hint that keeping people connected in contested places is a real use case for 150 Mbps-class links to ordinary phones.
Third, it sits next to the military network SpaceX is already building. In May the Space Force awarded SpaceX a $2.29 billion contract for the Space Data Network Backbone, the program once called MILNET. It’s a big constellation in low orbit for moving military data, and SpaceX owes a fully working prototype by the end of 2027. The broader Space Data Network supports Golden Dome missile defense and the Pentagon’s effort to link sensors and shooters across every branch (CJADC2). The Space Force also wants the Backbone to take over requirements planned for the Space Development Agency’s Tranche 3 data-relay satellites (DefenseScoop).
Our outlook (analysis, not reporting): the phone network runs on commercial airwaves and has no publicly announced military role. We found no reporting on a military-only (Starshield) version of phone-to-satellite service. Still, when one company runs the biggest commercial internet constellation, a military data backbone, and a phone network in orbit, defense planners will have to think hard about resilience. The near-term uses we’d expect are backup communications for bases and first responders, keeping allied civilians connected in a crisis, and tracking supplies and sensors where there’s no ground network. Commercial phone signals are easy to detect and locate, so this would sit alongside protected military satellite systems rather than replace them. The flip side is concentration: more of the country’s critical connectivity riding on one company’s network, prices and choices.
The carriers aren’t standing still
On Oct. 1, AT&T, T-Mobile and Verizon made their satellite joint venture official and named Paul Roth interim CEO. The idea is to pool their airwaves so satellite companies can plug in. Verizon CEO Dan Schulman doesn’t think satellites can take on carriers in cities and suburbs, where “95 to 98% of our revenues are,” calling satellite “a complementary service… not a direct competitor” (SDxCentral). T-Mobile’s CEO has put satellite-to-phone traffic at just 0.0002% of the company’s network use, and Light Reading reports that T-Mobile and Canada’s Rogers both removed Starlink branding from their satellite pages (Light Reading).
SpaceX has rivals in orbit, too. AST SpaceMobile wants to use spectrum owned by Ligado. Amazon Leo has filed for a phone-to-satellite system of more than 5,000 satellites and is trying to buy Globalstar. Lynk and Omnispace have merged into Elveo Mobile. The Ligado and Globalstar deals still need FCC approval (Broadband Breakfast; Light Reading).
Would people switch? It’s an open question. In a TD Cowen survey of 800 people, 34% said they’d consider a SpaceX phone service and 66% said they wouldn’t (Broadband Breakfast).
It all comes down to Starship (our math)
None of these deadlines work without Starship, SpaceX’s giant reusable rocket. SpaceX says Starship can carry “more than 50” V2 satellites per launch, that launches start in mid-2027, and that about 1,200 satellites would give “global and contiguous coverage” within six months (PCMag). Starship reached orbit for the first time on Sept. 28 and released 26 next-generation V3 Starlink internet satellites (Ars Technica; Reuters).
Here’s our back-of-the-envelope math, assuming 50 satellites per flight and a first launch in mid-2027:
| Goal | Satellites | Time window | Per year | Starship flights per year |
|---|---|---|---|---|
| First global coverage | ~1,200 | 6 months | ~2,400 (pace) | ~48 (about 24 flights in 6 months) |
| FCC’s halfway deadline | 7,500 | mid-2027 to Oct. 2032 (~5.3 yrs) | ~1,400 | ~28 |
| FCC’s final deadline | 15,000 | mid-2027 to Oct. 2035 (~8.3 yrs) | ~1,800 | ~36 |
| Second half only | 7,500 | Oct. 2032 to Oct. 2035 (3 yrs) | 2,500 | 50 |
Those numbers cover only the phone network. They don’t include replacing satellites that wear out and fall back at 330 km, and they don’t include everything else Starship has to do: V3 internet satellites, NASA’s Artemis Moon work, and commercial customers. For a sense of scale, SpaceX had launched nearly 13,000 Starlink satellites on its smaller Falcon 9 rocket by early October, according to astronomer Jonathan McDowell’s tracking as cited by TechSpot. SpaceX has shown it can build satellites at that pace. What it hasn’t shown yet is Starship flying several times a month.
How crowded is it up there, anyway?
Go back to September 2008, when SpaceX’s little Falcon 1 rocket first reached orbit. Space was a much quieter neighborhood. The Union of Concerned Scientists, which keeps a well-known database of working satellites, counted 905 active satellites in its update covering launches through Jan. 1, 2009 (UCS, via SeeSat-L). NASA’s Orbital Debris Program Office listed 12,743 cataloged objects in orbit on that same date (NASA ODQN, Jan. 2009). That figure covers everything the U.S. military’s tracking network logs: live satellites, dead ones, spent rocket stages and debris.
A decade later, just before SpaceX launched its first batch of 60 Starlinks in May 2019, UCS counted 2,062 active satellites as of March 31, 2019 (SpaceQ). NASA’s catalog had grown to 19,404 objects as of April 1, 2019 (NASA ODQN, May 2019).
Now look at today. As of Sept. 30, 2026, McDowell counts 16,913 active satellites in orbit, and 11,143 of them are Starlinks. He also counts 34,758 cataloged objects in orbit, including 11,166 pieces of cataloged debris (Jonathan’s Space Report). Our math: Starlink makes up about two of every three working satellites in orbit (roughly 66%).
| Moment | Active satellites | Cataloged objects in orbit |
|---|---|---|
| When SpaceX first reached orbit (data as of Jan. 1, 2009) | 905 (UCS) | 12,743 (NASA ODPO) |
| Just before Starlink’s first launch (spring 2019) | 2,062 (UCS, Mar. 31) | 19,404 (NASA ODPO, Apr. 1) |
| Today (Sept. 30, 2026) | 16,913, including 11,143 Starlinks (McDowell) | 34,758 (McDowell) |
These come from different counters with slightly different methods, so treat the table as a rough comparison. By our math, there are about 18 times as many working satellites as when SpaceX first reached orbit, about 8 times as many as just before Starlink began, and roughly 2.7 times as many cataloged objects as in early 2009.
The cataloged number is only what we can easily see and track. The European Space Agency’s 2026 Space Environment Report pulls in data from several tracking networks. It puts the regularly tracked population at about 47,140 objects and estimates there are 68,450 objects larger than 10 cm, about 1.5 million between 1 and 10 cm, and 230 million between 1 mm and 1 cm (ESA).
Three bad days that still haunt low orbit
Much of today’s debris traces back to a few events, and none of them involved Starlink:
- January 2007: China blew up its own old weather satellite, Fengyun-1C, with a missile at about 850 km. By May 2010, NASA had cataloged 2,841 fragments, with 2,756 still in orbit (NASA ODQN, 2010). Two years after the test, NASA noted that fewer than 2% of the cataloged pieces had fallen back and that many would stay up for decades (NASA ODQN, Jan. 2009).
- February 2009: A working Iridium satellite and a dead Russian satellite, Cosmos 2251, collided by accident at about 790 km. By May 2010, NASA had cataloged 1,267 Cosmos fragments and 521 Iridium fragments (NASA ODQN, 2010).
- November 2021: Russia destroyed its old Cosmos 1408 satellite with a missile. U.S. trackers spotted more than 1,500 trackable fragments, and 1,604 had been added to the catalog by March 7, 2022 (NASA ODQN, 2022).
McDowell’s count today still shows 2,656 pieces of ASAT (anti-satellite) debris and 770 pieces of collision debris in orbit (Jonathan’s Space Report).
And here’s what’s coming
These are numbers that have been authorized or planned, not launched:
- SpaceX Starlink Gen2: In January the FCC approved a second batch of 7,500, bringing authorized Gen2 to 15,000 of the 29,988 SpaceX asked for (SpaceNews; FCC DA 26-36). Its first-generation fleet was licensed separately.
- SpaceX phone network: 15,000, under this week’s order.
- SpaceX, longer term: Space.com reports SpaceX has applied to grow the constellation to 100,000 satellites (Space.com).
- Amazon Leo: 3,232 first-generation satellites, with 331 in orbit as of June. The FCC waived its halfway deadline but kept the July 30, 2029 deadline for the full fleet (Via Satellite).
- China’s Guowang: planned at about 13,000 satellites (SpaceNews).
- China’s Qianfan (“Thousand Sails”): planned at 14,000, as SpaceNews reported in 2025 (SpaceNews).
Our math: just those five line items (SpaceX Gen2, the new phone network, Amazon’s first generation, Guowang and Qianfan) add up to about 60,000 satellites. That’s roughly three and a half times everything working in orbit today, and it leaves out SpaceX’s first generation and its 100,000-satellite application.
Does that mean we’re headed for a pileup?
Here’s the honest answer: the risk is real, and so are the ways people are managing it.
The dodging is already constant. In its July 1, 2026 report to the FCC, SpaceX said Starlink satellites made 207,152 collision-avoidance maneuvers between Dec. 1, 2025 and May 31, 2026: 65,137 by first-generation satellites and 142,015 by Gen2. That works out to about 36 and 46 maneuvers per satellite per year. SpaceX moves when the odds of a collision pass 3 in 10 million, a bar it describes as about 100 times stricter than the industry norm of 1 in 10,000 (SpaceX semiannual report, via Scribd). Space.com tallied more than 355,000 maneuvers over the full year, more than three times the 2024 total (Space.com).
Experts are worried about where that curve goes. “I think we’re heading towards a situation where there will be a collision involving an operational satellite in the constellation,” the University of Birmingham’s Hugh Lewis told Space.com. “And it will not be for the lack of trying to avoid those things.” Tommaso Sgobba of the International Association for the Advancement of Space Safety explained why crowding gets worse faster than you’d think: “Double the satellites in a shell and you roughly quadruple the number of pairs that need to be watched.” Lewis says the safest answer is to keep constellations in separate altitude bands. He notes that China’s Thousand Sails is expected to occupy regions similar to Starlink’s (Space.com).
The nightmare scenario has a name. It’s called Kessler syndrome: collisions create debris that causes more collisions, in a chain reaction that can make an orbit unusable. ESA’s 2026 report warns that the debris population is growing again despite better cleanup practices, and that “the Kessler syndrome’s runaway effect” is “becoming visible much earlier in time.” It adds that even if no new satellites were launched, debris would keep increasing, because breakups create fragments faster than the atmosphere clears them (ESA).
Low orbit is the best tool we have. This is where the 330 km altitude starts to look smart. The higher you go, the thinner the air and the longer junk lingers. That’s why Fengyun-1C pieces at 850 km will stay up for decades. In January, SpaceX began moving about 4,400 Starlinks from 550 km down to 480 km. Nicolls said the move cuts the time a dead satellite would take to fall out of orbit during solar minimum, the quiet phase of the Sun’s 11-year cycle when the upper atmosphere thins out, from more than four years to just a few months (Ars Technica). When a Starlink suffered an in-orbit anomaly in December and shed hundreds of pieces of debris, trackers expected those pieces to decay within weeks, and SpaceX said the satellite itself would reenter within weeks too (SpaceNews). Our analysis: the new phone satellites will fly lower still, at 326–335 km, so anything that fails there should come down even faster. No one has published an exact figure for this fleet.
The rules have tightened. In 2022 the FCC adopted a “5-year rule” requiring low-orbit satellites to be disposed of within five years of finishing their missions, replacing the old 25-year guideline (FCC). This week’s order adds twice-yearly public reporting on close calls, maneuvers, retirements and reentries for the phone fleet.
Burning up isn’t free either. Low orbits keep space cleaner by sending satellites down to burn up, which raises a different question. A 2024 study in Geophysical Research Letters by Ferreira and colleagues estimated that a typical 250-kg satellite produces about 30 kg of aluminum oxide particles as it burns up. It put the 2022 total at about 17 metric tons and found that scenarios involving megaconstellations point to more than 360 metric tons a year, which “can lead to significant ozone depletion” (Ferreira et al., GRL 2024). In this order, the FCC called concerns about the atmospheric effects of reentering satellites “speculative” and declined to require an environmental review. Expect that debate to get louder as reentries climb.
The night sky is part of the bill. The FCC said optical astronomy “fall[s] outside of our review” and leaned on SpaceX’s promises to work with NSF and NASA. The National Radio Astronomy Observatory raised concerns about radio interference, DarkSky International asked for an environmental review, and the American Astronomical Society objected as well (TechSpot).
What could still go wrong
- Interference fights. DIRECTV, SES, Eutelsat/OneWeb and Viasat all objected to letting SpaceX exceed the limits that protect satellites in high geostationary orbit, the ones that park over one spot on Earth. The FCC allowed it inside the U.S. only, required SpaceX to steer clear of a 4-degree band around those satellites, and made clear SpaceX has to fix any harmful interference it causes. The Space Bureau can pull the waiver if problems aren’t resolved.
- Launch traffic. Blue Origin warned that a fleet this big could crowd out launch windows for everyone else. The FCC accepted SpaceX’s trajectory-screening tool and coordination promises as “currently sufficient.” The twice-yearly reports will show whether that holds.
- The business case. Carriers argue satellites can’t match towers for capacity in cities. SpaceX’s answer may include small ground radios called femtocells, an idea some in the industry have brushed off (PCMag).
- Too much in one basket. One company would control the biggest fleet in low orbit, nationwide S-band satellite rights, and a growing piece of military communications. Regulators so far see that as good for competition. Critics will push back.
What to watch next
- Oct. 29, 2026: The FCC votes on the new satellite-leasing rules (482 MHz) and the 25 MHz phone-to-satellite auction, which includes 2020–2025 MHz.
- Nov. 4, 2026: SpaceX’s surety bond is due. Its 30-day window to reject the conditions closes around the same time.
- Starship’s pace: the next flights from Starbase in Texas and the first launches from LC-39A in Florida.
- Mid-to-late 2027: the first V2 Starlink Mobile launches.
- Around Nov. 30, 2027: the final step of the EchoStar deal closes, which unlocks commercial service on the new airwaves.
- Federal coordination: agreements covering the 27 protection zones, and whether any defense user signs up.
- Carrier moves: what T-Mobile decides about Starlink, which satellite partner the carrier venture picks first, and whether any wholesale deal for SpaceX to resell a carrier’s network emerges.
- The first twice-yearly report on the phone fleet: close calls, maneuvers and reentries.
The bottom line
The FCC has given SpaceX the license, the airwaves and the independence to build a phone network in orbit. What’s left is physics, rocket cadence, competition, and the shared job of keeping low Earth orbit usable as it fills up. If Starship delivers, ordinary phones getting real broadband from space could feel normal well before the 2032 deadline. For the rancher on that dirt road, that’s simple: a call that goes through. For carriers, the Pentagon and everyone who cares about a clean, safe orbit, it means a lot of new planning.
Reported facts come from the FCC order and the sources linked throughout. Anything labeled “our read,” “our analysis,” “our outlook” or “our math” is Redstone Orbital’s own analysis. The ranch, boat, fire-crew and flood scenes are illustrations, not real people or events.
Sources
-
FCC
SpaceX approved for NGSO MSS system for direct-to-device services (DA 26-1078, GN Docket 25-340)
-
FCC
Authorization and Order, DA 26-1078 (full text)
-
FCC
Order Granting SpaceX-EchoStar Applications (DA 26-471)
-
Light Reading
FCC OKs SpaceX to deploy 15K D2D satellites, and nine years to complete it
-
Via Satellite
FCC Grants SpaceX Approval for 15,000 Direct-to-Device Satellites in MSS Spectrum
-
PCMag
FCC Green Lights SpaceX's 15,000 Satellites for Starlink Mobile
-
Broadband Breakfast
FCC Approves SpaceX Request for 15,000 Direct-to-Device Satellites
-
TechSpot
FCC approves 15,000 SpaceX Starlink satellites that could compete directly with cellular carriers
-
SpaceNews
FCC approves SpaceX spectrum deal with $2.4 billion escrow condition
-
PCMag
SpaceX Makes 'Starlink Mobile' Official, But It's Not Competing With Carriers
-
TelecomTV
MWC26: Starlink preps new D2D constellation
-
Teslarati
SpaceX to launch Starlink V2 satellites on Starship starting 2027
-
Ars Technica
SpaceX's Starship goes orbital, deploying first next-gen Starlinks
-
Reuters
SpaceX's Starship makes orbital debut deploying Starlinks before early ending
-
SDxCentral
AT&T, T-Mobile US, Verizon-led satellite JV is official, gains temporary leader
-
DefenseScoop
Space Force awards $2.29B deal to SpaceX to accelerate 'backbone' SATCOM network
-
CNBC / Reuters
Pentagon spars with SpaceX over Starlink price hike during Iran war
-
Reuters
US lets Starlink provide direct-to-cell coverage for hurricane-hit areas
-
Union of Concerned Scientists (via SeeSat-L)
UCS Satellite Database update: 905 active satellites (launches through Jan. 1, 2009)
-
SpaceQ
Number of Active Satellites Now Over 2,000 and Has Doubled in the Last Six Years
-
NASA ODPO
Orbital Debris Quarterly News, Vol. 13 Issue 1 (Fengyun-1C two years later; box score Jan. 1, 2009)
-
NASA ODPO
Orbital Debris Quarterly News, Vol. 14 Issue 3 (Top ten satellite breakups)
-
NASA ODPO
Orbital Debris Quarterly News, Vol. 23 Issues 1 & 2 (box score Apr. 1, 2019)
-
NASA ODPO
Orbital Debris Quarterly News, Vol. 26 Issue 1 (Cosmos 1408 breakup)
-
Jonathan's Space Report (Jonathan McDowell)
Active satellites and cataloged objects in orbit (data as of Sept. 30, 2026)
-
ESA Space Debris Office
Space Environment Statistics (DISCOSweb)
-
ESA
ESA Space Environment Report 2026
-
SpaceNews
FCC approves 7,500 additional Starlink satellites
-
FCC
SpaceX Gen2 Upgrade Applications Partial Grant (DA 26-36)
-
Via Satellite
FCC Gives Amazon Leo 50% Deployment Waiver, With Conditions on Spectrum Priority
-
SpaceNews
China launches Guowang, Yaogan-40 satellites, sets up busy Q4 manifest
-
SpaceNews
China launches fourth batch of Thousand Sails megaconstellation satellites
-
SpaceX (via Scribd)
SpaceX Gen1-Gen2 Semi-Annual Report to the FCC (July 1, 2026)
-
Space.com
Every SpaceX Starlink satellite has to dodge a collision almost weekly, and experts fear the worst
-
Ars Technica
SpaceX begins "significant reconfiguration" of Starlink satellite constellation
-
SpaceNews
SpaceX to lower orbits of some Starlink satellites
-
FCC
FCC Adopts New '5-Year Rule' for Deorbiting Satellites (FCC 22-74)
-
Geophysical Research Letters
Potential Ozone Depletion From Satellite Demise During Atmospheric Reentry in the Era of Mega-Constellations (Ferreira et al.)