Every nation's digital economy rests on a surprisingly thin stack of physical infrastructure: a handful of subsea cable landing stations, a few major internet exchange points, and terrestrial fibre routes that often follow the same river valleys and highway corridors. A single cable cut, a targeted cyberattack, or a natural disaster can sever a country's external connectivity for days or weeks. Nations that rely exclusively on commercial satellite operators for backup face queue prioritisation, foreign-government pressure on those operators, and service terms that can be suspended during exactly the crises when the link is most needed.
A sovereign backup internet constellation changes the equation. A LEO constellation of Ka-band or V-band microsatellites, operated from national ground infrastructure, provides burst-capable broadband that automatically activates when terrestrial routes degrade below a threshold. The satellite layer does not need to match the full capacity of a nation's peacetime internet; it needs to carry essential government services, financial clearing, emergency broadcast, and enough public-facing capacity to prevent societal disruption. Fifty to eighty satellites in a walker constellation can deliver that for a mid-sized nation with latency under 30 ms.
The operational outcome is a resilience floor the government controls end-to-end. Traffic routing decisions, encryption standards, priority queuing, and lawful-intercept compliance all remain under national jurisdiction. The system doubles as a sovereign testbed for domestic satellite manufacturing and spectrum coordination, building industrial capacity that compounds over successive generations of the constellation.
Frequently asked
Why can't we just contract Starlink or OneWeb as a backup instead of building our own?
Commercial providers like Starlink (SpaceX) and OneWeb (Eutelsat) operate under the licensing and export-control jurisdiction of their home countries. A foreign government or regulator can compel service suspension — as events in the 2022 Ukraine conflict illustrated when access decisions became geopolitically contested. Owning the satellites means owning the kill switch. A sovereign system also keeps traffic routing, encryption keys, and data residency under national law.
What is the minimum viable constellation size for a national backup?
For a mid-latitude nation with a land area up to roughly 500,000 km², a LEO constellation of 12–18 microsatellites at 500–600 km altitude can provide adequate revisit for data bursting and continuous coverage when augmented by inter-satellite links. Smaller island states may achieve adequate coverage with as few as 6–8 satellites in a well-designed orbital plane. ESA and national space agencies can model this precisely against traffic requirements and ground-station placement.
How does satellite backup interoperate with existing terrestrial networks during a partial outage?
The architecture typically uses software-defined routing at internet exchange points (IXPs) that automatically fail over to satellite uplinks when terrestrial BGP routes become unreachable. Ground stations feed into the national backbone at multiple geographically diverse injection points. Standards such as ETSI EN 302 307-2 (DVB-S2X) and open-standard ground-segment software (e.g. OpenSAND) enable interoperability with existing IP infrastructure without proprietary lock-in.
Who owns the orbital slots and spectrum licences in a sovereign system?
The ITU assigns spectrum and orbital positions to national administrations (not to companies) through the ITU Radio Regulations filing process. A nation that files its own coordination under Article 9/11 of the Radio Regulations retains those rights permanently, subject to the due-diligence and bring-into-use rules. Leasing spectrum from a foreign operator means those rights revert to the filing administration — a critical sovereignty gap.
What cybersecurity standards apply to the satellite link and ground segment?
The ground segment should be hardened to at least NIST SP 800-53 Rev. 5 control baselines (or the national equivalent) for the classified/government traffic tier. The space-to-ground link should implement authenticated command uplinks per CCSDS 352.0-B-2 (Space Data Link Security Protocol) to prevent command spoofing or hijacking. End-to-end encryption of user traffic should comply with national cryptographic standards and avoid dependence on foreign-controlled key management infrastructure.
How long does it take to procure and launch a sovereign nanosatellite constellation?
A realistic timeline for a greenfield programme — from mission requirements through design, build, test, launch, and operational acceptance — is 4–7 years for a first-generation constellation of 12–20 satellites. Nations with existing space agencies and established industrial partners (e.g. prime contractors vetted through ESA's ECSS procurement standards) can compress this to 3–4 years. Spectrum coordination runs in parallel and is often the critical path.
Can a small or low-income country afford this?
A 12-satellite LEO microsatellite backup constellation can be built for $80–200 M USD depending on the technology tier selected, with annual operations costs of $10–20 M. The World Bank estimates a single prolonged national internet outage costs lower-middle-income countries $24–68 M per day in GDP; the constellation pays back its capital cost within months of preventing one major event. Multilateral financing through the World Bank Digital Development Partnership or regional development banks is an established route.
What happens to the backup system if the primary terrestrial cables are cut for months, not hours?
A backup satellite system designed for short outages will be overwhelmed if terrestrial cables are out for weeks or months (as happened to Tonga after the 2022 Hunga Tonga volcanic eruption, which severed the only subsea cable). A resilient design requires higher-capacity satellites, pre-positioned terminal stock for civilian distribution, and traffic management policies that prioritise government and critical-infrastructure users. The system design brief must specify both the peak outage duration and the sustained throughput requirement — not just a headline 'backup' capability.