Musk's Plan to Kill Dead Zones: SpaceX's Rooftop Mini Cell Tower Beams Satellite Internet Indoors
Elon Musk says SpaceX has finished designing the missing piece of its plan to put a phone signal everywhere on Earth — a miniature cell tower that sits on your roof.
Speaking at the New York Times DealBook Summit on November 18, Musk said the rooftop microcell would solve the one problem satellites cannot: getting a signal indoors. "We've got a plan that enables ubiquitous access on Earth, including via cell phone," he said — but a metal or thick roof blocks the signal from space. "So we need something that's terrestrial that addresses situations where people are, you know, in a building with a thick roof essentially."
40x more bandwidth is coming
SpaceX will launch its Direct-to-Cell V2 satellite constellation next year, which Musk says will increase bandwidth 40-fold. The rooftop units are the ground half of the plan: a small base station you mount on a roof "like a Starlink dish," in Musk's words, that carries the signal the last few metres indoors. He expects production to begin in 2027 — though he demurred on the timing, saying the devices were "going into production, I think, next year." Musk has a well-documented history of missing his own production deadlines.
A $25 billion spectrum bet behind it
The hardware sits on a spectrum war chest SpaceX has spent two years assembling. It won FCC approval in 2024 to connect Starlink directly to phones, in partnership with T-Mobile, struck a $17 billion deal with EchoStar in 2025, and agreed this year to acquire a further portfolio of low-band spectrum valued at $8 billion. SpaceX president Gwynne Shotwell said on the August earnings call that the company would skip big cell towers entirely and deploy small, low-power base stations instead. One option Musk floated is bundling the mini towers with Starlink terminals.
The plan has fuelled predictions that legacy carriers are in danger. Musk does not buy that. "I don't think this means, like you know, AT&T, Verizon, or others go out of business or anything," he said — but they may lose some revenue. His argument is that demand for bandwidth is about to explode anyway, driven not by people but by machines: "a lot, and that bandwidth will be driven primarily by superintelligence and robots, and they're going to need a lot more bandwidth than humans."
Source: TechCrunch (November 19, 2026).
Comments
Post a Comment