What exactly is direct-to-device satellite connectivity, how does it differ from satellite phones and satellite broadband, and what can it do today?
The short answer. direct-to-device (D2D) means an ordinary mobile phone, or a phone with a standard modem chip, connecting straight to a satellite without a dish, an external antenna or a separate satellite handset. In 2026 it mostly delivers text messages, emergency alerts, location sharing and light data in places with no mobile signal, at a cost of nothing to about $10 a month to the user. It is a coverage product, not a capacity product: it fills the gaps in mobile networks rather than replacing them. Three technical routes exist, each with different spectrum, phones and economics, and the competition between them, and over the spectrum they use, is the subject of this study.
Three kinds of satellite connectivity
| Type | User equipment | Examples | Typical use |
|---|---|---|---|
| Satellite broadband | Dish or flat-panel terminal | Starlink, Amazon Leo, OneWeb | Homes, businesses, ships, aircraft: tens to hundreds of Mbps |
| Mobile satellite services (MSS) with dedicated handsets | Satellite phone or tracker | Iridium, Globalstar, Inmarsat, Thuraya | Voice, messaging and data in remote areas; maritime; government |
| Direct-to-device | An ordinary smartphone or a phone with a satellite-capable modem | Starlink Direct to Cell (T-Mobile T-Satellite and others), AST SpaceMobile, Apple–Globalstar, Skylo, Iridium NTN Direct, Huawei–Tiantong | Messaging, emergency, light data where there is no mobile coverage |
The distinction matters because the economics differ by orders of magnitude. A satellite broadband terminal has an antenna that gives tens of decibels of gain; a smartphone antenna has almost none (Chapter 4). That single fact makes D2D far harder and far lower in capacity per satellite than broadband.
Exhibit 2.1
The same satellites can serve dishes and phones, but the economics differ by orders of magnitude
Satellite broadband versus direct-to-device
| Broadband to a dish | Direct to an ordinary phone | |
|---|---|---|
| Antenna on the ground | ~0.5 m² dish, high gain | Phone antenna, ~0 dBi |
| Typical throughput | Hundreds of Mbps | Kilobits to a few Mbps |
| Price | $50–120 a month | $0–10 a month |
| Use | Primary connection | Occasional, outdoors, emergencies |
| Business | Capacity | Coverage |
Three technical routes
| Route | How it works | Phones | Spectrum | Main players |
|---|---|---|---|---|
| Satellite as a cell tower in mobile bands | The satellite transmits a standard LTE or 5G signal in a mobile operator's own frequencies; the phone thinks it is a distant tower | Unmodified phones (most LTE phones) | Terrestrial mobile spectrum, leased from or shared with the operator | Starlink Direct to Cell, AST SpaceMobile, Lynk (now Elveo) |
| Satellite in dedicated satellite bands with a standard chip | The phone's modem supports the 3GPP non-terrestrial network (NTN) standard and talks to satellites in MSS bands | New phones with NTN-capable modems (e.g., Pixel 9 and later, Galaxy S25) | MSS spectrum (L-band, S-band, 2 GHz) | Skylo (using partner satellites), Iridium NTN Direct, Viasat/Inmarsat, Equatys |
| Proprietary integration | The phone maker integrates a specific satellite system in hardware and software | Specific phone models (iPhone 14 and later; Huawei Mate 60 Pro and later) | MSS spectrum owned by the partner | Apple–Globalstar (moving to Amazon), Huawei–Tiantong |
Sources. company information; Chapters 10 to 20.
Each route has a different trade-off. The first reaches existing phones but uses scarce terrestrial spectrum and must avoid interfering with ground networks. The second needs new phones but uses satellite spectrum designed for the purpose and follows a global standard. The third is fast to market but ties a phone maker to one satellite partner.
What D2D can do in 2026
| Service | Status, October 2026 | Notes |
|---|---|---|
| Emergency alerts (wireless emergency alerts) | Live (Starlink with partners; used in disasters since 2024) | Broadcast to all compatible phones in an area |
| Emergency messaging (SOS) | Live (Apple since 2022; Skylo on Android since 2024; Starlink with partners) | Usually free to the user |
| Two-way text messaging | Live (T-Mobile from July 2025; Verizon via Skylo from March 2025; others) | Seconds to minutes per message in early systems |
| Light data for selected apps | Live (T-Satellite from October 2025 for maps, messaging, weather and others) | A few apps optimised for low bandwidth |
| Voice calls | Limited; AST demonstrated; Starlink and others planned | Needs much more capacity per user |
| Broadband-like data | Promised for next-generation systems (Starlink's 15,000-satellite system; AST Block 2) | Depends on spectrum and very large arrays |
Sources. operator and company announcements.
How much it is used
Early usage is small. T-Mobile's chief executive said that, on May 2026 data, satellite traffic was about 0.0002% of the operator's total network usage, mostly in national parks, and lower than the company had expected. SpaceX's filing says its mobile service reaches about 7.4 million monthly unique devices across about 30 countries. Those two facts are consistent: many devices touch the satellite system occasionally, for a few messages, and almost none use it for sustained data. That is what a coverage product looks like.
Why it matters anyway
| Reason | Explanation |
|---|---|
| Coverage | Mobile networks cover most people but a minority of land and sea; D2D covers everywhere a phone can see the sky |
| Safety | Emergency messaging and alerts where there is no signal; disaster response when towers fail (Chapter 34) |
| Competition among operators | Satellite coverage is becoming a feature operators advertise and bundle |
| Platform control | Phone makers (Apple, Google, Samsung, Huawei) and satellite owners (SpaceX, Amazon) compete over who owns the customer relationship |
| Spectrum value | D2D has made satellite and mobile spectrum interchangeable at the margin, driving multibillion-dollar spectrum deals (Chapter 26) |
Why the definition matters for analysis
Precise definitions matter because the three kinds of satellite connectivity have very different economics. Satellite broadband to dishes is a capacity business with tens of millions of potential subscribers paying $50–120 a month. Dedicated satellite phones and terminals are a niche business with a few million users paying for reliability. D2D to ordinary phones is a coverage business with potentially billions of devices, most of which will use it rarely and pay little or nothing directly.
Confusing them produces errors. Applying broadband revenue per user to D2D's device counts produces implausibly large markets. Applying satellite-phone pricing to D2D's audience ignores that ordinary phone users are unwilling to pay those prices. This study keeps the categories separate and, where it estimates revenue, distinguishes paying users from devices that connect (Chapter 7).
A note on terminology
The industry uses several names for the same idea. "Direct-to-device" and "direct-to-cell" (SpaceX's term) refer to satellites connecting to unmodified or lightly modified phones. "Supplemental Coverage from Space" is the US regulatory term for satellites using operators' terrestrial spectrum. "Non-terrestrial networks" is the 3GPP standards term, which also covers high-altitude platforms. "Satellite-to-phone" and "space-based cellular broadband" (AST's term) are also used. This study uses D2D throughout except when quoting.
What would change this view
A next-generation system delivering reliable voice and tens of megabits per second to ordinary phones, at scale, would turn D2D from a coverage product into a capacity product that competes with terrestrial networks in rural areas. Continued low usage despite availability would keep it a low-revenue safety feature bundled into existing plans.