Direct-to-Device Satellite Service: Architecture and Business Case

The idea of connecting an ordinary smartphone directly to a satellite (no specialized terminal, no external antenna) has been an aspiration of the satellite industry for decades. The combination of massive LEO constellation deployments, improved satellite EIRP (Effective Isotropic Radiated Power), and 3GPP standardization of Non-Terrestrial Networks (NTN) in Release 17 has brought this vision to commercial reality. As of early 2026, Apple's Emergency SOS via satellite, T-Mobile's partnership with SpaceX, and AST SpaceMobile's commercial launch have each demonstrated different points on the D2D capability spectrum.

The Link Budget Challenge

The fundamental physics challenge for direct-to-device service is the enormous link budget deficit between a LEO satellite and a smartphone. A standard smartphone antenna has a gain of roughly 0–3 dBi and maximum transmit power of 200–250 mW (23 dBm). A LEO satellite at 500–600 km altitude introduces approximately 160–165 dB of free-space path loss at typical cellular frequencies (600–2100 MHz).

The math is unforgiving: to close the link to a standard handset, satellites must either be extremely large (to concentrate beam gain), use very low data rates (to reduce required SNR), or accept very high latency architectures. These constraints drive the three distinct technical approaches currently deployed:

• Narrowband / store-and-forward messaging: Texting, SOS alerts, and low-bandwidth IoT data. Link budget is achievable with smaller satellites because the narrow channel bandwidth improves noise floor. Apple/Globalstar and the Garmin inReach approach fall here.
• NB-IoT / LTE-M via NTN: 3GPP Release 17 defined NTN operation for NB-IoT and LTE-M, enabling licensed-spectrum connectivity from LEO satellites using existing cellular standards. Throughput is limited (tens of kbps) but sufficient for voice, SMS, and basic data alerts. T-Mobile/SpaceX operate in this tier.
• Full broadband via large arrays: AST SpaceMobile's BlueBird satellites use very large deployable phased arrays (roughly 64 m² on commercial satellites) to produce sufficient EIRP for broadband links to unmodified handsets. This approach is technically elegant but requires the largest and most expensive satellites.

Spectrum Considerations

D2D services operate in licensed terrestrial spectrum: typically mobile network operator (MNO) spectrum in the 600–700 MHz, 850 MHz, AWS, or 2.1 GHz bands. This is deliberate: smartphones are already certified to operate on these bands, so no device modification is needed. The satellite must be licensed (through the MNO's regulatory position or a direct satellite license) to transmit on these frequencies.

Operating in terrestrial spectrum raises interference considerations. When a satellite beam illuminates a large geographic area (a major technical constraint: large beams reduce gain), it may interact with terrestrial base stations using the same frequencies. Sophisticated interference coordination mechanisms, frequency planning, and time-division approaches are required to maintain compatibility with existing terrestrial networks. This is an active area of regulatory work at the FCC and ITU, with evolving rules on protection criteria and coordination procedures.

Business Model Variants

The business model for D2D varies significantly by capability tier:

• Emergency/SOS tier: Typically offered as a device feature (by handset OEMs like Apple) or as a carrier add-on. Revenue model is device licensing fees to the satellite operator and potential carrier per-use fees. Not expected to be a large standalone revenue stream.
• Supplemental coverage tier: MNOs pay satellite operators for coverage extension in areas where terrestrial network deployment is uneconomical. The satellite operator provides wholesale capacity; the MNO retails it to subscribers under existing plans. T-Mobile/SpaceX beta service operates on this model.
• Broadband D2D: If full broadband can be delivered to unmodified handsets globally, the total addressable market includes the several billion people with cellular-capable devices in areas with inadequate terrestrial coverage. AST SpaceMobile's commercial case targets this opportunity, with MNO partnerships in multiple regions.

Technical Constraints and Limitations

Even with large satellites and favorable link budgets, D2D service has real constraints that differentiate it from terrestrial cellular:

• Throughput: Per-user throughput in early systems is limited, suitable for messaging and voice, not high-definition streaming. The total capacity of a single large satellite shared across many users in a beam is far lower than a terrestrial macro cell.
• Latency: Round-trip latency from LEO to device and back is 20–40 ms for the propagation component alone, comparable to terrestrial LTE but with additional processing delay. Voice and basic data work fine; real-time gaming does not.
• Doppler: A LEO satellite moves at roughly 7.5 km/s relative to the ground, producing significant Doppler shift that must be compensated. NTN standards include Doppler pre-compensation at the satellite to minimize handset complexity.
• Handoff: Continuous connectivity requires seamless handoff between satellites as they pass overhead, which requires sophisticated beam management and inter-satellite coordination.

Outlook

Direct-to-device satellite service is one of the most technically interesting developments in the satellite industry in a decade. The combination of commercial scale, standardized air interfaces (3GPP NTN), and global smartphone penetration creates a potentially large market. Near-term growth will be driven by supplemental coverage agreements with MNOs; longer-term, full broadband D2D has transformative potential for underserved markets.

Track satellite launches and constellation deployments relevant to D2D networks at SpaceNexus Launch Tracker and Satellite Tracker.