2
From the newsletter (sign-up if you want it sooner): Amazon’s planned 5,105-satellite LEO constellation signals a shift in the D2D race – from proving satellite connectivity to embedding it within mainstream telco networks. The real challenge is no longer coverage, but sustainable economics, seamless integration, and scalable commercial services.
Sure, Amazon’s announcement that it will launch 5,105 satellites into low-Earth orbit to create a D2D constellation from 2028 is another salvo in the battle with Starlink, AST SpaceMobile, and others. But it is also a line in the sand to deploy a business model that works. The race to connect regular smartphones from orbit, or just to prove it can be done, is done. The contest is to integrate non-terrestrial networks (NTNs) into mainstream telco infrastructure, with a business-model wraparound to put D2D into the hands of consumers and enterprises.
Amazon seems to get this. Its engineering sounds good – on-board signal processing, laser inter-satellite links, beam hopping, adaptive modulation. But almost every paragraph in its blog about it emphasises partnerships with mobile operators – and not competition against them. It is treading carefully. Vodafone, Herotel, Australia’s NBN, and Apple are presented as collaborators in a broad telco ecosystem, where it wants to be – rather than as customers of a standalone satellite network, with nefarious plans to eat their lunch (and all of that).
As Sean Kinney writes, the industry has moved beyond asking whether D2D is technically feasible. The challenge is industrialisation: manufacturing, launching, and operating enough satellite capacity at sustainable economics to support millions of users with predictable performance. Much of the D2D debate – to date – has focused on coverage. That is becoming the easy part. Performance, orchestration and economics are much harder. D2D services are valuable because they replace ‘no signal’ with ‘some signal’. They don’t replicate terrestrial 5G.
Consumers may value emergency messaging and out-back connectivity, but convincing them to pay a monthly premium is a tough commercial ask. Enterprises place higher value on resilience than bandwidth. Utilities, transport operators, logistics companies, defence organisations, and public safety agencies care less about peak throughput than guaranteed connectivity when terrestrial infrastructure disappears. Amazon’s messaging hints at this transition. Its examples go well beyond smartphones into fleet management, disaster recovery, remote industrial ops and IoT.
Which are all less about dead zones, and more about critical comms. The industry’s vocabulary is already changing from D2D to NTN, and increasingly from NTN towards 3D networks. Instead of treating terrestrial and satellite as separate technologies, telcos are viewing them like complementary access layers within a single comms platform. Fiber, 5G, Wi-Fi, and LEO networks become different routes in the same system rather than competitive propositions. Amazon Leo looks / sounds like it is designed to slot into existing mobile systems.
Amazon’s pending $11.57 billion acquisition of Globalstar brings spectrum rights, satellite assets, and key telco relationships that reduce friction. Its emphasis on standards, spectrum coordination, and compatibility sound like a firm that wants to play ball – and integrate, not disrupt. It also aligns with where 3GPP standards are heading. Successive 3GPP releases are steadily expanding NTN capabilities; early 6G thinking assumes convergence is the way. AI orchestration, policy control, and digital twins will be as important as the radio tech – if telco systems are to switch between.
The irony is that satellites may become less visible just as they become more important. If the industry succeeds, users will stop thinking about satellite connectivity altogether. Their device will simply remain connected, whether through fiber, public/private 5G, public/enterprise Wi-Fi, or satellites in orbit. Amazon’s announcement is mostly about preparing for that future, rather than just the volume of launches. The winners won’t be the ones with the biggest constellations, necessarily, but those that make satellites disappear into the network.
