Clusters of tiny satellites flying in coordinated formation could, in theory, deliver stronger or cheaper cellular coverage to ordinary smartphones than relying on a smaller number of large satellites. That proposition, advanced in a recent techno-economic analysis of global non-terrestrial networks, challenges the prevailing industry bet on flagship hardware and reframes the direct-to-phone competition around fleet architecture rather than individual spacecraft size. The idea arrives as early commercial tests of satellite-to-handset links have already proven the basic technology works, raising a pointed question: what orbital design actually scales?
From Lab Demos to Live Calls
The technical foundation for beaming cellular service from orbit to unmodified phones has been validated in stages. AST SpaceMobile’s BlueWalker 3 test satellite was described in a company SEC filing as reaching key regulatory and testing milestones around April 25, 2023, after the company reported completing two-way voice calls to standard smartphones. Those tests also confirmed initial compatibility through SIM and network information exchange, meaning the satellite could authenticate with existing carrier infrastructure without requiring special handsets.
Separately, per the European Space Agency, ESA partnered with Telesat and Amarisoft to achieve what the agency called a world-first 5G 3GPP non-terrestrial network link over LEO, using a 3GPP Release 17-compliant stack and a Telesat LEO satellite. That demonstration showed standards-based cellular protocols could traverse a low-Earth-orbit path end to end, not just in simulation but through actual satellite hardware.
These milestones matter because they prove the radio-frequency link budget closes: a phone’s low-power transmitter can reach orbit, and a satellite’s signal can reach a phone on the ground. The open question is no longer whether direct-to-phone works but how to make it affordable and dense enough to serve billions of potential users.
Starlink’s Beta and What Crowdsourced Data Reveals
SpaceX has moved fastest toward commercial scale. Starlink, in partnership with T-Mobile, has been cited in an arXiv preprint as running broad beta trials of so-called Supplemental Coverage from Space, with plans to support voice and data services by mid-2025. That timeline, if met, would give Starlink a head start over competitors still assembling their constellations and negotiating spectrum deals.
An independent crowdsourced measurement study of Starlink’s direct-to-device radio access network offers a reality check on how the service actually performs at the handset layer. By collecting data from real users rather than relying on company-reported benchmarks, the research captures coverage and availability patterns that marketing materials tend to smooth over. Variable signal quality and intermittent connectivity appear as expected features of early satellite-to-phone service, not bugs, given the geometry of fast-moving LEO satellites passing over fixed ground positions.
This kind of independent verification is critical. Carrier partnerships and press releases describe what a system is designed to do. Crowdsourced measurements describe what it actually does. The gap between those two accounts will shape regulatory confidence, investor expectations, and consumer adoption as satellite-to-phone moves from novelty to infrastructure.
The Swarm Alternative
Most current direct-to-phone architectures rely on relatively large, expensive satellites carrying oversized antenna arrays. BlueWalker 3, for instance, deployed a phased-array antenna spanning roughly the area of a small apartment. That approach concentrates capability in single points of failure and drives up per-unit launch and manufacturing costs, even if it simplifies some aspects of network control.
A recent techno-economic framework hosted on arXiv proposes a different path: distributing the same aggregate capability across many smaller, cheaper pico-satellites flying in coordinated swarms. The model connects constellation architecture to coverage economics, exploring how swarms of tiny spacecraft could enable different coverage and capacity profiles compared to fewer large platforms carrying equivalent total mass.
The logic is straightforward. A swarm of dozens of pico-satellites can spread its coverage footprint more evenly across a region, dynamically reallocating beams to match where users actually are rather than illuminating large fixed cells. If one unit fails, the swarm degrades gracefully instead of losing an entire coverage zone. And because each satellite is simpler, production can scale on shorter timelines with lower capital risk per unit, which matters in a market where standards and spectrum policy are still evolving.
Conference proceedings from the AIAA/USU Conference on Small Satellites reinforce this line of thinking. Research on “palmsat” pico-satellite mission scenarios, according to that conference paper, found that pico-satellites offer a relatively cost-effective way to demonstrate new technologies for Earth observation or space science. Whether that kind of cost advantage translates to telecommunications is less certain, but proponents argue it could shorten iteration cycles and encourage more experimental architectures.
Beam Coordination as the Hard Problem
Flying a swarm is easy compared to making it act as a single coherent antenna. For pico-satellite clusters to match or exceed the signal quality of a large phased array, the individual units must synchronize their transmissions with extreme precision. That requires tight timing, accurate knowledge of relative positions, and fast coordination links among satellites that may be separated by kilometers.
In fractionated spacecraft concepts, engineers often introduce a central “master” node that handles tasks such as time distribution, navigation updates, and global resource allocation. Applied to non-terrestrial networks, a master satellite or ground segment controller could assign frequency bands, schedule which pico-satellites illuminate which users, and adjust power levels to avoid self-interference as the swarm sweeps over populated regions.
The payoff is potentially transformative. If a swarm can steer many narrow beams instead of a few broad ones, it can reuse spectrum more aggressively, packing more simultaneous connections into the same orbital footprint. That, in turn, could bring satellite-to-phone capacity closer to what users expect from terrestrial 4G and 5G networks, rather than treating space-based coverage as a last-resort emergency layer.
Who Builds and Validates the Models
The debate over large satellites versus swarms is not playing out only in corporate R&D labs. Much of the underlying analysis appears first in open-access preprints that the broader community can scrutinize. The direct-to-device measurement work and the swarm-focused techno-economic study both sit on arXiv, a repository supported by a network of institutional members ranging from universities to research labs.
That funding model matters for a fast-moving field like non-terrestrial networks. Rather than waiting for long journal review cycles, engineers and policy analysts can read, critique, and build on new ideas as soon as they are posted. arXiv, in turn, relies on a mix of member support and individual contributions, inviting researchers and readers alike to help sustain the service so that emerging work on satellite connectivity and other topics remains openly accessible.
For practitioners trying to interpret these preprints, arXiv also offers detailed guidance for users that explains how submissions are moderated, categorized, and linked to later journal versions when available. That context is useful when techno-economic proposals for satellite swarms intersect with commercial roadmaps and regulatory filings, where the incentives and language can differ sharply.
Economics, Risk, and the Road Ahead
Ultimately, the choice between a few big satellites and many small ones is not purely technical. It is an economic and risk-management decision. Large spacecraft can deliver powerful beams and may simplify some aspects of network design, but they concentrate capital in hardware that is costly to replace and slow to iterate. Swarms distribute both capability and risk, but demand sophisticated coordination and may face more complex regulatory scrutiny around spectrum use and collision avoidance.
The early evidence from direct-to-phone trials suggests that users will tolerate some intermittency in exchange for coverage where terrestrial towers are absent. What remains unclear is how quickly expectations will rise once satellite links are marketed as extensions of mainstream mobile plans rather than emergency lifelines. If demand for higher throughput and lower latency grows, architectures that can scale capacity through mass-produced pico-satellites may look more attractive than monolithic platforms that are difficult to upgrade.
For now, the industry is effectively running a live experiment in orbit. Starlink’s large satellites, AST SpaceMobile’s apartment-sized arrays, and the swarm concepts emerging from academic and open-access modeling each embody different bets on how best to blanket the planet in signal. As more measurement campaigns, techno-economic analyses, and regulatory decisions accumulate, the question of “what orbital design actually scales” will move from theory to market verdict.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
