End of the cell dead zone: how satellites are coming for your phone

Everyone has a dead-zone story. The stretch of motorway where the maps app froze. The cottage where the one-bar signal meant standing in the garden. The valley where the emergency number just spun. For a century of telephony, the answer was always the same: another mast, another trench of fibre, another decade of waiting.

This autumn, the answer changed — or at least, the official version of it did. In the space of one week, the US Federal Communications Commission cleared SpaceX to use $19.6 billion worth of airwaves for phones that talk straight to satellites, scheduled a vote on auctioning off more spectrum for the same purpose, and watched America’s three biggest mobile carriers formalise a joint venture whose stated mission is to “nearly eliminate dead zones”. Direct-to-device — the industry’s unlovely name for a phone that needs no mast at all — is no longer a demo. It is a regulatory, commercial and military fact.

Flip News has spent the year reporting each step of this shift as it happened. This feature pulls those threads together: what was decided this week in Washington, how the technology actually works, where in the world it is landing first, why governments are simultaneously begging for it and fighting it, and what could still stop it.

The week Washington decided the dead zone was over

The headline development came from the FCC. Late on Tuesday 6 October, the commission approved SpaceX’s plan to use the 65 megahertz of spectrum it bought from EchoStar — two deals announced last year, worth a combined $19.6 billion — to expand its direct-to-device plans. A day later, the commission announced it will vote on 29 October on whether to advance a new spectrum auction aimed at direct-to-device services.

FCC chair Brendan Carr laid out the ambition in unusually concrete terms. The commission is proposing to auction 25 megahertz of key spectrum for D2D and to make an additional 482 megahertz available for supplemental coverage from space. Carr said in September that upcoming spectrum auctions could generate more than $100 billion. “The ‘direct-to-device’ proposals work to continue leveraging this cutting-edge tech to end cell phone dead zones and provide service directly from next-gen satellite constellations to your smartphone,” Carr said in a statement.

The backdrop matters. The 15,000-satellite authorisation — the January approval of 7,500 additional second-generation Starlink satellites — stands as the underlying constellation grant. What changed this week was the airwaves to talk to phones through them: spectrum bought for mobile use is what turns a broadband constellation into, potentially, a mobile network.

That possibility is exactly what alarmed America’s incumbents. On 1 October, AT&T, T-Mobile and Verizon finalised a joint venture agreement — first sketched in May — to pool spectrum resources for direct-to-device satellite use. Wireless industry veteran Paul Roth was named interim chief executive while the search for a permanent boss runs, with a board drawn from each carrier overseeing operations. The venture aims to expand coverage in underserved areas, provide backup connectivity during natural disasters, and build a unified, standards-based approach to D2D consumer services. No name, ownership split or financial terms have been disclosed.

Some analysts see the venture as defensive, Reuters reported — and with reason. SpaceX president Gwynne Shotwell said in August that the company will build the infrastructure needed to make Starlink “a true mobile service” and expects to win “quite a few” customers from the big three carriers. Existing arrangements stay in place: T-Mobile keeps its commercial D2D service with SpaceX, Verizon offers satellite messaging with Skylo, and AT&T and Verizon separately work with AST SpaceMobile.

The carriers are not exactly thrilled about the competition. AT&T chief executive John Stankey called SpaceX’s mobile plan “not a viable strategy” in an Axios interview at the end of September, attacking the economics: “It would cost as much to do that as it would to build a macro network to be able to handle those capabilities.” He added a legal objection — “You can’t radiate cellular signals from somebody’s house without their permission, unless somebody changes the law on that” — and insisted that for home broadband, fibre wins: “Satellite, at its best — if everything happens right over the next 10 years — is still not going to beat fiber.” Stankey is the second US carrier boss to reject the plan this autumn; the Times of India reported that T-Mobile chief executive Peter Osvaldik said earlier in September the strategy should be “laughed out of the room”, arguing satellite signals cannot penetrate buildings, walls or car windscreens.

Note the shape of the argument, though. Nobody in the industry now disputes that satellites will serve phones in dead zones — the fight is over who controls the service, who gets paid, and whether it complements terrestrial networks or replaces them.

How direct-to-device actually works

The idea is disarmingly simple. Your phone already talks to masts miles away; a satellite in low Earth orbit is a mast that never ends, moving at 27,000 kilometres per hour, and your phone just has to catch it as it passes. The technical trick is doing it with an ordinary phone, at the frequencies phones already use, with no dish, no antenna and no extra equipment.

The first generation is already live and modest. In Kazakhstan, Beeline and Starlink Mobile launched direct-to-cell service at the end of September with the support of the country’s Ministry of Artificial Intelligence and Digital Development. Compatible smartphones connect directly to Starlink satellites; the connection establishes itself automatically when terrestrial coverage disappears, and at the initial stage users can send and receive messages and access basic internet services where no terrestrial network exists. The ministry was explicit that the service is designed to complement, not replace, ground networks — for remote regions, highways, mountainous areas, and places where ground infrastructure has been damaged.

The second generation is where the dead zone genuinely ends. Japan’s KDDI has signed an Asia-first agreement with SpaceX for Starlink Mobile V2, to be offered through its au brand. The V2 satellites carry larger antennas delivering roughly 20 times the throughput per satellite of the first generation and, with more communications beams, about 100 times the data density — enough for native voice calls including VoLTE, streaming, web browsing and high-speed apps on compatible smartphones. “We are grateful to KDDI for trusting us to deliver resilient connectivity to their customers with our next-gen V2 technology,” said Gwynne Shotwell. KDDI aims to start the service by the end of 2027 in areas beyond terrestrial coverage — mountainous regions and remote islands — upgrading the au Starlink Direct service it launched in April 2025, which offers messaging and supported app data. The detail that makes the point: Japan’s mobile networks already cover more than 99.9% of the population, and even there, satellite-to-phone is the answer for the last fraction of a percent.

The map is filling in

Starlink’s footprint is expanding faster than any single story can capture, and the week’s most striking new datapoint came from India. Starlink’s Lauren Dreyer told India Mobile Congress on Wednesday that the company has built its India operations specifically to meet the country’s security and regulatory requirements — “establishing controls at every layer for India, and keeping Indian user data within India” — with 20 gateway sites comprising hundreds of antennas, and is now awaiting the government’s go-ahead to launch. “Starlink offers the biggest potential to help accelerate the speed to connect a 100 per cent of India, and people can get connected within minutes, not years,” she said. The Centre has granted satcom licences to three companies — Starlink, Bharti Airtel and Reliance Jio — but spectrum and security clearance are still needed before services begin. Starlink has partnered with both Airtel and Jio.

Elsewhere, the subscriber numbers are becoming real. In Kenya, the Communications Authority reported that Starlink Internet Services Kenya recorded 27,616 fixed subscriptions in the quarter to June 2026 — about 99.7% of all satellite internet connections in the country, in a fixed broadband market where fibre still dominates but satellite subscriptions grew 54.4% year on year.

Bosnia and Herzegovina is weeks away: communications minister Edin Forto said Starlink will begin operating “in a few days or weeks”, after meeting Starlink’s director for global licensing and calling the service “serious infrastructure that can replace what BiH has not done for the development of broadband high-speed Internet”. “We cannot move forward as a country without infrastructure that will enable companies and citizens to have access to such an Internet,” Forto said. A local entity, “Starlink BH” d.o.o., is registered, and a licence application has been with the regulator since May.

The holdout that says most about the politics of all this is South Africa. The country is absent from the 150-plus nations where the FCC cleared Starlink Mobile for global roaming in September — while all six of its neighbours appear. The obstacle is the rule requiring telecoms licensees to be 30% owned by Black citizens, which Elon Musk has called racist. Deputy communications minister Mondli Gungubele has now made a public appeal: “Can you get along with us? Work with us in this journey of minimising inequality?” he asked, adding: “I hope Elon is listening.” On the ownership rule itself, he argued: “What is 30% share if you take into account the 70% value that you are going to make in terms of yields and profits? He could invest very significantly in the country of his birth.” Talks, he said, continue: “We’re engaging on all those issues, and we hope to find one another.” SpaceX wants the regulator to accept equity-equivalent investment programmes instead of ownership stakes — it has offered free internet to 5,000 rural schools — but the regulator says the law itself would need amending.

The pattern is consistent across every market: regulators hold the keys, and Starlink’s expansion is a negotiation — over data sovereignty in India, local ownership in South Africa, licensing in Bosnia — not a conquest.

The same signal is a different kind of weapon

The dead-zone story and the war-zone story are the same story. A network that reaches every phone on Earth without a mast also reaches every soldier, every protester and every missile launcher.

In Ukraine, that duality is the entire argument. Finland’s President Alexander Stubb told Reuters in New York on 24 September that he had pleaded with Elon Musk to expand Starlink coverage so Ukraine can strike Russian ballistic missile launchers inside Russian territory. “I plead that Elon Musk would give the Ukrainians the possibility to use Starlink in areas where there are ballistic missile launchers, either in the occupied territory of Ukraine or in Russia for strategic purposes because that would save lives,” he said. Ukraine’s presidential office chief Kyrylo Budanov has pressed the same case harder: “We are extremely interested in the capabilities provided by Starlink / Starshield. We urgently need these capabilities to repel attacks and strike Russian missile systems from which missiles are launched that terrorise Ukraine’s civilian population.” He stressed: “We need enhanced capabilities — we need wider coverage,” saying Ukraine’s current access is insufficient. Ukraine currently has Starlink access only on its own territory; SpaceX has restricted the network’s operation inside Russia, and Kyiv wants connectivity to support long-range drones targeting launchers, command posts and military infrastructure up to 200 kilometres from the border.

The Kremlin’s answer was blunt. Spokesman Dmitry Peskov warned that providing Starlink systems for strikes deep inside Russia would be “very detrimental, potentially dangerous and reckless”.

And the ground infrastructure itself is now a target. On the night of 23 September, a fire broke out at the Starlink ground station at Wola Krobowska, about 40 kilometres south of Warsaw — a facility owned by state telecom Exatel that routes traffic including to Ukraine. Firefighters found a bottle containing a flammable substance near the damaged equipment. Deputy prime minister Krzysztof Gawkowski called it sabotage: “This act of sabotage was deliberately designed to disable the station … disrupting internet access for various institutions, including the Ukrainian army,” adding there was “much to suggest that we are dealing with an element of the next phase of hybrid warfare.” Prime minister Donald Tusk said he would wait for “100% certainty” before using the word himself, while conceding the incident “does not look good”. The station stayed operational.

Other governments are reacting to the strategic dependency, not the coverage. Canada’s prime minister Mark Carney told the New York Times in September that Canada will “diversify away from” Starlink as part of cutting strategic dependence on the United States, citing the “extreme tail risk” of US action after President Trump’s repeated rhetoric about making Canada the 51st state. “I think you have a responsibility in these roles to look at extreme tail risk,” Carney said. “That’s just risk management.” He acknowledged no rival network yet matches Starlink, and suggested working with Europe and potentially South Korea; Ottawa has already committed a $2.14-billion loan to Telesat’s Lightspeed network. The Canadian turn has history: Ontario cancelled a $100-million Starlink contract in March 2025 over Trump’s tariffs, later settling confidentially for a “kill fee” said to be far less than the contract’s value.

In Iran, the satellite is the enemy of the state. Deputy communications minister Mohammad Hafez Hakami warned that growing use of Starlink could deal the “final nail in the coffin” to Iran’s ability to control cyberspace, admitting the shutdown strategy that escalated from platform blocks in 2017 to a nationwide blackout in 2019 is becoming “less effective”. “If this continues, in the next crisis we may be forced to cut electricity,” he said. Tehran has since seized hundreds of terminals and arrested users. Iran summoned Norway’s ambassador on 4 October over Starlink’s “unauthorised” operations, accusing Oslo of failing its duties as the registering administration for the Starlink network at the International Telecommunication Union — and then claimed it had “successfully disabled” Starlink terminals during the January protests, “for the first time”, saying it has “other tools” to counter the service.

Moscow, meanwhile, is building its own. Russia’s Rassvet network, developed by Bureau 1440, launched its first 16 low-orbit satellites in March with a second batch in July, and aims for operations in 2027 — 292 satellites by 2027 and 924 by 2035 against more than 10,000 Starlinks. Ukraine has urged Washington to sanction the firms behind it: Kyiv’s acting ambassador Denys Sienik warned that “if they launch Rassvet by the end of 2026, they will be able to target to hit Ukrainian targets real time on our territory”, and that without sanctions the programme would be a “game-changer”. The issue is now personal between presidents: Volodymyr Zelenskyy told the Financial Times he has asked Donald Trump “several times” to sanction the Russian and Chinese companies involved, and the US administration faces an 18 October deadline under the sanctions law Trump signed on 18 September. Former Russian president Dmitry Medvedev has warned that blocking Rassvet would trigger “full-scale space warfare”.

The challengers are not waiting

Starlink’s lead is real, but the week also showed how seriously others take the threat. Britain’s Open Cosmos, backed by the government’s defence venture capital fund, plans to launch almost 200 satellites by 2030 — 96 by 2028, growing to 192 — to give Europe a sovereign alternative to Starlink. Orbiting 650 miles up, the network would offer resilient communications for governments and companies when traditional telecoms infrastructure fails or becomes overloaded. “Right now, there is a big need in Europe to avoid dependency on SpaceX and on Amazon and on systems that are from abroad,” founder Rafel Jordà Siquier told The Times. The company unveiled its ConnectedCosmos concept at MWC Barcelona in March, holds Ka-band spectrum rights, and plans to manufacture up to 200 satellites a year from its Harwell campus.

On the military side, Boeing’s Insitu and aviation Wi-Fi provider Gogo have successfully tested a military drone using satellite communications — a modified Galileo HDX antenna on Insitu’s Integrator aircraft, linked through Eutelsat OneWeb rather than Starlink. The system would be ITAR-free, easing exports. “Adding an ITAR-free, high-speed SATCOM option is an essential capability for our global customers,” said Insitu’s John Kelly. The test lands as European governments and parts of the Pentagon voice concern about over-reliance on Starlink and its Starshield government service.

Underwriting all of this is launch cadence nobody else can match. SpaceX flew its 80th Starlink mission of 2026 on 6 September — 27 V2 Mini satellites from Vandenberg on a booster flying its 19th mission — and is shifting Florida Starlink launches to Starship. At the end of September, Starship reached orbit for the first time on Flight 14, deploying 26 next-generation V3 satellites before an engine failure cut the mission short and it splashed down early off Hawaii. Each V3 supports about 1 terabit per second of downlink — roughly ten times the current V2 satellites — and 26 of them carry a combined 26 terabits per second of capacity. The V3 design underpins the Gen 2 satellites for Starlink Mobile and the direct-to-device service that will use the EchoStar spectrum. That is the physical reason the FCC’s decision matters now: the satellites that can actually use those airwaves are going up.

What comes next

Three dates structure the rest of the year. On 29 October, the FCC votes on whether to advance the new direct-to-device spectrum auction — the procedural step that determines how fast the airwaves Carr is offering actually reach the market. On 18 October, the Trump administration must decide which Russian entities to sanction under the law signed in September, with Rassvet’s component suppliers at the top of Ukraine’s list. And by the end of 2027, if KDDI’s plan holds, Japan gets native voice calls from space — the first mass-market proof that a phone can be a satellite terminal.

What does it mean for the reader? First, the obvious: the dead zone is being attacked from three directions at once — regulators releasing spectrum, carriers building a shared D2D venture, and satellites with an order of magnitude more capacity. Second, the caveats that matter: spectrum auctions take time, rural capacity is still the binding constraint until the V3 constellation scales, and every government along the way gets a veto over licensing — South Africa’s empty slot on the roaming map is the reminder that coverage is negotiated, not automatic. Third, the uncomfortable truth: the same infrastructure that lets your phone work in a valley lets a drone work in a war zone, and the geopolitics of orbit are now a permanent part of the story.

For a hundred years, the mobile dead zone was a fact of geography. This week, Washington, the carriers and the launch pads agreed it is a policy problem instead — one with a schedule, a spectrum price tag and a launch manifest. The geography is not giving up without a fight. But for the first time, the fight has a date: 29 October.

Sources

Flip News reporting synthesised in this feature (full articles in site/content/satellite/):

• 2026-10-07-2045-fcc-spacex-echostar-spectrum-d2d-approval.md — FCC approves EchoStar spectrum use, 29 October auction vote, Carr’s $100bn claim, 15,000-satellite authorisation
• 2026-10-01-2152-carriers-d2d-spectrum-jv.md and 2026-10-03-0246-big-three-carriers-satellite-jv-roth.md — AT&T/T-Mobile/Verizon joint venture, Paul Roth interim CEO
• 2026-10-01-0232-att-stankey-starlink-phone-plan.md — Stankey’s “not a viable strategy”, Osvaldik’s “laughed out of the room”, Shotwell’s “true mobile service”
• 2026-09-18-1710-kddi-au-starlink-mobile-v2-japan.md — KDDI Starlink Mobile V2, 20x throughput, end-2027 target
• 2026-09-28-1610-starlink-mobile-launches-kazakhstan-beeline.md — Beeline Kazakhstan direct-to-cell launch, first in Central Asia
• 2026-10-07-1455-starlink-india-ready-20-gateways-awaiting-go-ahead.md — Starlink India, 20 gateways, Dreyer’s India Mobile Congress remarks
• 2026-09-25-1640-starlink-kenya-27600-subscribers.md — Kenya regulator data, 27,616 subscriptions, 99.7% of satellite connections
• 2026-09-29-1330-starlink-bosnia-launch-weeks.md — Forto’s “few days or weeks” announcement
• 2026-09-29-1736-south-africa-urges-musk-starlink-30-percent.md and 2026-09-23-1251-starlink-mobile-south-africa-misses-out.md — Gungubele’s appeal, the 30% rule, the roaming-map gap
• 2026-09-25-0815-poland-starlink-ground-station-fire-sabotage-probe.md — Wola Krobowska fire, Gawkowski’s sabotage claim
• 2026-09-25-0256-stubb-pleads-musk-starlink-ukraine-strikes.md — Stubb’s plea to Musk
• 2026-09-27-2158-budanov-wider-starlink-starshield-patriot.md — Budanov on Starlink/Starshield coverage
• 2026-09-27-2258-kremlin-starlink-strikes-deep-russia-dangerous.md — Peskov’s “dangerous and reckless” warning
• 2026-09-25-0400-carney-canada-starlink-diversify-away.md — Carney’s diversification pledge, Ontario contract cancellation
• 2026-10-01-0715-iran-starlink-internet-control-warning.md — Hakami’s “final nail in the coffin” warning
• 2026-10-05-1640-iran-summons-norway-envoy-starlink-crackdown-claim.md — Norway envoy summoning, terminal-disablement claim
• 2026-10-01-0255-ukraine-urges-us-sanctions-russia-starlink-rival.md and 2026-10-03-1228-zelenskyy-asks-trump-sanctions-rassvet-starlink-rival.md — Rassvet sanctions push, Sienik, Zelenskyy’s “several times”
• 2026-10-04-1805-medvedev-warns-space-warfare-rassvet-blockade.md — Medvedev’s space-warfare warning
• 2026-10-04-2210-open-cosmos-200-satellite-starlink-rival.md — Open Cosmos 200-satellite plan
• 2026-09-23-1650-boeing-insitu-gogo-military-drone-starlink-alternative.md — Insitu/Gogo OneWeb drone test
• starlink-80th-mission-2026.md — 80th Starlink launch of 2026, Vandenberg, B1088
• starship-flight-14-v3-starlink.md and 2026-09-29-0245-starship-flight-14-splashdown.md — Flight 14 first orbital flight, 26 V3 satellites, 1 Tbit/s downlink
• Named external sources as cited in those articles: Reuters, The Hindu BusinessLine, The Fast Mode, Qazinform, Sarajevo Times, Moneyweb/Bloomberg, MyBroadband, AFP, Gulf News/Iran International, Rudaw, The Times (UK), Times of India, Axios/Inc.