Satellite IoT in 2026: When Your SIM Connects to a Satellite Instead of a Tower — and What That Actually Costs
June 4, 2026 · 7 min read · Technical Whitepapers
Iridium's NTN NB-IoT service launches commercially in 2026. Starlink Direct-to-Cell has 1.8M+ subscribers on T-Mobile for text messaging. AST SpaceMobile is flying 6 satellites the size of basketball courts. The satellite IoT market is projected from $1.82B to $15.77B by 2035. This is not a future technology — it is a procurement decision that needs to be made now for devices deploying beyond cellular range in 2027.
Satellite IoT is not one thing. It is three architectures competing for different device classes: Direct-to-Cell (Starlink, AST SpaceMobile) connects unmodified phones via terrestrial spectrum — messaging today, broadband eventually. 3GPP NTN NB-IoT (Iridium, Skylo) connects IoT sensors via globally harmonized MSS spectrum — commercial launch in 2026. Proprietary satellite IoT (Iridium Edge, Globalstar) has been shipping for years — low-data-rate telemetry for assets beyond cellular range. The procurement decision is not "satellite or cellular" — it is "which satellite layer for which device class, at what airtime cost, with what fallback to terrestrial."
The Three Architectures
| Architecture | Providers | Devices | Data Rate | Latency | Power | Cost/Typical Use |
|---|---|---|---|---|---|---|
| ------------- | ----------- | --------- | ----------- | --------- | ------- | ----------------- |
| Direct-to-Cell (D2C) | Starlink, AST SpaceMobile | Unmodified phones/tablets | Text today, voice/data ~2027+ | Seconds to minutes | Phone-class (daily charge) | SMS: included in MNO plan; broadband: TBD |
| 3GPP NTN NB-IoT | Iridium (Project Stardust), Skylo | NTN-capable NB-IoT modules (Nordic, Qualcomm) | 2-40 kbps | 30-60 seconds per message | Ultra-low (PSM-compatible) | ~$3-8/month for 100-byte/hour telemetry |
| Proprietary satellite IoT | Iridium Edge, Globalstar, Orbcomm | Proprietary modules | 0.3-2.4 kbps | Minutes | Low (primary battery) | $5-15/month for daily location ping |
The key difference for IoT procurement: NTN NB-IoT uses the same 3GPP protocol stack as terrestrial NB-IoT. The device firmware does not need to know whether it is talking to a tower or a satellite. The SIM, the module, and the network handle the routing. This is the architecture that makes hybrid satellite-cellular viable at scale — one SKU, one SIM, one firmware build, two physical layers.
Iridium NTN NB-IoT — The IoT Standard That Launches in 2026
Iridium's Project Stardust adopted 3GPP NTN NB-IoT as its IoT standard in 2024. MNO trials began in summer 2025. Full commercial launch is 2026. Deutsche Telekom is integrating Iridium NTN for NB-IoT direct-to-device in 2026. Nordic Semiconductor is building Iridium NTN Direct into its modules.
The February 2026 Iridium 9604 module integrates satellite + cellular + GNSS into a single platform — 60% board space reduction versus discrete modules. It supports location-based network selection: cellular when available, satellite when not. For a soil sensor in rural Brazil, a container tracker in the South Atlantic, or a pipeline monitor in the Canadian Arctic, this means one module, one SIM, one data plan — with the physical layer switching transparently.
The cost: Iridium NTN NB-IoT airtime is projected at $3-8/month for a device transmitting 100 bytes/hour. Compare to Iridium Edge (proprietary) at $5-15/month for a daily location ping. NTN NB-IoT is cheaper per byte and more power-efficient because it uses the same PSM/eDRX mechanisms as terrestrial NB-IoT. The satellite is just another cell tower — with a 30-60 second latency penalty.
Source: The Mobile Network, "Iridium pushes on with satellite NB-IoT plans, targets 2026 launch", May 2025. Available at https://the-mobile-network.com/2025/05/iridium-pushes-on-with-satellite-nb-iot-plans-targets-2026-launch/
Starlink Direct-to-Cell — Messaging for Emergency, Not IoT Telemetry
Starlink D2C uses T-Mobile's terrestrial spectrum to deliver text messaging to unmodified phones via Starlink satellites. 1.8M+ subscribers as of early 2026. Partners include Rogers, Optus, One NZ, Salt, Kyivstar, Entel, KDDI. The use case is emergency messaging and SMS in areas without cellular coverage — not IoT sensor telemetry.
The limitation for IoT: Starlink D2C operates in terrestrial MNO spectrum, requiring country-by-country regulatory approval. It is designed for consumer phones, not power-optimized IoT modules. It does not support NB-IoT PSM/eDRX. For a sensor that wakes once per hour and transmits 100 bytes, Starlink D2C is the wrong tool — too power-hungry, too expensive per byte, and not designed for machine-to-machine traffic. For a vehicle emergency call button or a remote worker safety check-in, it is the right tool.
When Satellite IoT Pays for Itself
The procurement crossover: a cellular-only device in a location with no coverage generates zero data and requires a site visit to retrieve. A hybrid satellite-cellular device in the same location transmits data daily and never requires a site visit. The satellite airtime cost ($3-15/month) is compared against the amortized cost of a site visit ($200-500 per visit, including labor, travel, and downtime). If the device would require more than one site visit per year without satellite connectivity, satellite pays for itself.
For agriculture: a 500-hectare farm with 50 soil sensors beyond cellular range. Without satellite: data collected manually once per season. With satellite: daily data, irrigation optimized, yield improved 10-25%. Satellite airtime for 50 sensors at $5/month = $3,000/year. Yield improvement on $500,000 of crop output at 10% = $50,000. The ROI math is not in the connectivity cost — it is in the data that the connectivity enables.
Source: Bitget News / Iridium, "Iridium 9604: Wagering on the Hybrid IoT S-Curve", February 2026. Available at https://www.bitget.com/news/detail/12560605228528