New frequencies for Starlink: multi billion dollar resources for direct to cell

SpaceX has purchased an additional portion of spectrum from the company EchoStar for $17 billion. An asset that, with smartphones connecting directly to satellites, will become increasingly valuable.

BY EMILIO COZZI

SpaceX purchased the rights to use an additional portion of the electromagnetic spectrum (a total of 50 MHz) in the S band in the United States, along with global licenses for Mobile Satellite Service (Mss). Elon Musk’s space company paid $17 billion.
The news is highly specialized, unlikely to make front page headlines. Yet it represents a massive investment, a sign of how power in the satellite communications market is becoming increasingly concentrated.

The licenses acquired belonged to the American company EchoStar Corporation, specialized in satellite communications. EchoStar had been developing its own constellation for direct to cell (or direct to device, D2D) service, meaning data and internet traffic to mobile devices (smartphones and tablets, the Internet of Things, cars, ships, and planes), not via ground antennas but from satellites orbiting the Earth.
In other words, EchoStar had been a potential competitor to Starlink, strategically focused on D2D to expand its user base and service offering, including 5G.

The past tense is necessary because, apparently, EchoStar will no longer be building a constellation, at least not in the near future. The hardware, about one hundred satellites, that it was supposed to launch into low Earth orbit to activate the service was the subject of a contract with Canada’s MDA Space (the company that built the Canadarm for the International Space Station), announced in August 2025 and valued at $1.3 billion.

Shortly after the sale of frequencies to SpaceX (and to AT&T, more on this shortly), Mda itself announced the termination of the contract as “the result of a sudden change in EchoStar’s strategy and business plan, following discussions on spectrum allocation with the Federal Communications Commission (Fcc) in the United States. EchoStar agreed to sell the AWS4 and H block spectrum licenses to SpaceX.” The deal involves $8.5 billion in cash and the same amount in SpaceX shares. As part of the agreement, Boost Mobile customers, EchoStar’s terrestrial wireless service, will gain access to Starlink’s direct to cell services.
Before Musk’s company stepped in, another US telecom giant had already seized on the carcass of the struggling competitor. At the end of August, for $23 billion in cash, AT&T announced the acquisition of about 30 MHz of mid band spectrum nationwide at 3.45 GHz, and about 20 MHz of low band spectrum nationwide at 600 MHz, frequencies for terrestrial antenna connections. EchoStar chose this path to resolve a dispute with the Fcc.

The company had in fact been under investigation since May, following complaints from SpaceX and others alleging that EchoStar was failing to meet the conditions for implementing its 5G network and leaving valuable frequencies unused, frequencies that now belong to the two giants. At the close of the transaction, the Fcc ended the investigation.

Starlink and the Others

In the satellite communications market, another major player is expected to make its debut: Kuiper, branded Amazon, is exploring solutions for D2D technology.
AT&T, however, is also in the game thanks to its agreement with AST SpaceMobile to build a constellation of about 250 low Earth orbit satellites for broadband service (so far, only five have been launched). Nothing comparable to the mega constellations of thousands of nodes, but enough to complement an already powerful terrestrial network, especially across the United States.
It should be noted that AT&T, with its “classic” connection service based on ground antennas and fiber, boasts a user base that far exceeds even SpaceX’s: more than 100 million compared to Starlink’s 6 million.
It is worth noting, however, that the latter is aiming for an architecture based solely on space based repeaters, capable of providing global coverage. It is active in more than 100 countries and promises speeds exceeding 100 megabits per second. These are far from the performances guaranteed by fiber, but competitive with wireless connections. Above all, SpaceX and Starlink are the only ones, for now, able to target multiple markets simultaneously: the institutional and public market (research agencies, governments, militaries), providing services to populations, activities, and contingents in areas not covered by ground antennas; and the private market, from airline and shipping fleets to individual smartphones, ensuring service continuity even in emergencies.

A Precious Asset

Frequencies cost billions because, like water, oil, or land, they are a limited resource. As already seen in Italy, competition for their acquisition and regulation (in 2024 there was a clash with TIM precisely on this point) could become fiercer in every single country.
Once again, it must be stressed that Starlink is ten steps ahead of everyone else and has already deployed direct to cell service through its partnership with T Mobile. The scale of investment shows how direct to device is considered the future of communications, often at the expense, for example, of radio astronomy. Scientists have long been warning about an increasingly deafening noise.
Recent developments confirm that their protests are not prevailing. And it is unlikely that anything will change in the near future, at least for them.