National and regional LEO broadband

A regional LEO broadband constellation gives a government direct control over connectivity infrastructure, spectrum rights, and the terms of access. This page sets out what that actually costs, how long it takes, and where the limits are.

The dependence this ends: Starlink and OneWeb dependence

The dependence you are currently running on

Starlink and OneWeb are foreign-licensed, foreign-operated systems. Their terms of service are set in Washington and London respectively. Spectrum filings are held by their parent companies under ITU coordination procedures that give those companies, not your government, the right to modify, suspend or terminate service in your territory. Ukraine discovered in 2022 that Starlink coverage decisions over active conflict zones are made unilaterally by the operator. The United Kingdom, as OneWeb's majority shareholder post-rescue, has explicit national-security override provisions baked into that company's governance. If your country's critical connectivity runs on either system, a foreign boardroom or foreign ministry sits between your citizens and their internet.

The post-2022 environment has sharpened this. Cable sabotage in the Baltic and elsewhere has demonstrated that terrestrial fallback is not always available. IRIS2, the European Union's sovereign LEO broadband programme, was formally initiated precisely because member states concluded that dependence on non-European constellations was a strategic liability. The arithmetic is uncomfortable: building your own constellation costs more in capital than buying capacity from an existing operator. The question is whether that premium buys something worth having. For most governments, the answer depends on what services run over the link and what happens when it goes dark.

What a regional constellation actually is, and is not

A global broadband constellation like Starlink requires thousands of satellites to maintain continuous coverage everywhere. A regional constellation does not. If your coverage requirement is a defined geographic area, perhaps a nation-state, an exclusive economic zone, or a regional bloc, the orbital geometry changes substantially. A constellation covering a region the size of Southeast Asia or sub-Saharan Africa can achieve continuous coverage with tens to low hundreds of satellites in inclined low Earth orbits, depending on the latitude band, minimum elevation angle required, and acceptable inter-beam handover frequency. The IRIS2 programme, covering European territory and selected partner regions, is sized at 290 satellites. That is a useful public reference point for a continental-scale ambition.

User terminals are the part of this discussion that gets skipped most often. LEO broadband requires electronically steered phased-array terminals that track satellites moving across the sky at roughly 7.5 kilometres per second. These are not simple dishes. They are manufactured at volume by a small number of suppliers, and their cost, currently in the low hundreds of dollars per unit at commercial scale, is falling but has not yet reached the price point of a mobile handset. Any honest programme plan accounts for terminal procurement, customs, distribution logistics, and the support infrastructure to keep them working in the field. The satellite is the visible part. The terminal network is where programmes succeed or fail on the ground.

Inter-satellite optical links change the architecture significantly. Without them, every satellite must hand off traffic to a ground station within its footprint, which limits how useful the constellation is over ocean or territory without ground infrastructure. With optical inter-satellite links, traffic can be routed across the constellation and down through a small number of gateway stations. This reduces ground infrastructure cost but adds per-satellite mass and complexity, and optical link availability is affected by orbital geometry and, to a lesser degree, atmospheric conditions at the ground terminal. It is not a free upgrade. It is a design trade-off that must be made early, because it affects bus sizing, power budgets, and launch mass.

The ambition ladder: constellation to full sovereign programme

There is no useful pathfinder for LEO broadband in the way there is for Earth observation or navigation augmentation. A single demonstration satellite proves orbital mechanics and some payload function, but it does not deliver connectivity to users. The minimum useful entry point is a constellation that achieves continuous regional coverage, and that means committing to a multi-satellite programme from the outset.

At constellation scale, covering a single nation or sub-region with continuous service, public precedents suggest programmes in this class require satellite counts in the range of twelve to sixty, depending on orbital altitude between roughly 500 and 1,200 kilometres, inclination chosen to match the coverage zone, and whether inter-satellite links are included. Publicly reported budgets for small-satellite LEO communications constellations of this scale, drawing on disclosed figures from national and commercial programmes, sit in the range of several hundred million to low billions of dollars over a five-to-eight-year development and deployment period. IRIS2's published programme budget of 10.6 billion euros covers a 290-satellite constellation with full ground segment and user-terminal ecosystem across a continent; that figure provides a ceiling reference, not a floor.

A full sovereign programme adds the ground segment, the national gateway infrastructure, the spectrum filing and coordination process at the ITU, the mission control centre, operator training, and the legal framework for domestic service licensing. These are not afterthoughts. ITU spectrum coordination for a new LEO constellation can take three to seven years if filings are contested, and the filing must be submitted before the first satellite launches to protect priority rights. Governments that have not already initiated ITU filings are already behind any competitor that has.

The honest arithmetic of build versus dependence

Buying capacity from Starlink or OneWeb is cheaper in year one. That is simply true. The capital cost of a sovereign constellation is front-loaded, the operational cost is ongoing, and the break-even point against leased capacity depends heavily on how much bandwidth you need and for how long. For a government whose connectivity requirement is modest and whose political risk tolerance is high, leasing may remain rational. The sibling page on sovereign gateways for leased constellations covers that option without pretending it does not exist.

The case for building shifts when three conditions align: the services running over the link are genuinely critical to national security or economic continuity; the government has no credible legal mechanism to compel the foreign operator to maintain service on its terms; and the country has, or can develop, the institutional capacity to operate the system. On the third point, capacity is not inherited. It is built through training, through staged handover, and through sustained investment in a national team. Programmes that skip this step end up with a constellation they cannot operate without the original contractor, which is a different kind of dependence.

What you own at the end, and what the limits are

A completed sovereign LEO broadband programme transfers to the customer: the satellites themselves and all associated intellectual property agreed under source-access terms; the ground station infrastructure including gateway and mission control facilities; the ITU spectrum filing and coordination record in the country's name; trained national operators capable of running day-to-day mission control; and the service licensing framework for domestic users. These are real assets with real balance-sheet value and real operational obligations.

The limits are equally real. A regional constellation of this scale cannot compete with Starlink on raw capacity or on terminal price at global volumes. Latency from LEO is genuinely low, typically under 40 milliseconds round-trip, which is a legitimate advantage over GEO. But throughput per user is a function of how many users share each beam, and a smaller constellation with fewer beams has less total capacity than a larger one. Satellite lifetimes in LEO at 500 to 1,200 kilometres are typically five to seven years before atmospheric drag and radiation degradation require replacement; a replenishment plan is part of the programme, not an optional extra. And if the ITU filing is not in place before launch, the frequency rights are not protected. There is no workaround for that.

What this mission is built from

Inclined low Earth orbits: Defines the orbital shell and inclination that matches the target coverage latitude band, balancing continuous regional coverage against satellite count.
Constellation geometry and revisit design: Determines how many orbital planes and satellites per plane are needed to achieve the required coverage continuity and minimum elevation angle over the service region.
Small satellite platforms (150–500 kg): Provides the bus platform sized to carry a communications payload and, where specified, optical inter-satellite link terminals within the mass and power budget of a shared or dedicated launch.
Communications payloads: The Ka-band or Ku-band phased-array payload that generates the spot beams serving user terminals and gateway links.
Optical inter-satellite links: Enables on-orbit routing between satellites, reducing the number of ground gateways required and extending useful coverage over ocean and ungoverned territory.
Ka-band high-rate stations: Ground gateway infrastructure that aggregates traffic from the constellation and connects it to the national internet exchange or government backbone.
Spectrum and ITU filings: Manages ITU coordination filings in the customer's name, establishing legal frequency priority rights before the first satellite is launched.

What you end up owning

The satellite constellation, with hardware audit rights and source-access terms agreed before contract signature
Ground gateway and mission control facilities, built and located on national territory
ITU spectrum filing and coordination record, held in the customer government's name
Trained national operations team capable of independent day-to-day mission control
Service licensing framework and the legal right to wholesale or retail connectivity domestically
Full documentation package: interface control documents, flight software source code under agreed terms, and replenishment specifications

Handover is staged across the programme: national engineers join mission control from first launch, assume primary operator status by mid-constellation deployment, and hold full independent operational authority before the contract closes. Satellize retains no ongoing operational role after handover, though launch and integration partners for replenishment satellites are arranged on terms the customer controls. Source-access and audit rights are contractually fixed before signature, not negotiated after delivery.

Programme parameters

Constellation size (regional coverage)	12 to 60 satellites depending on coverage zone, altitude and inter-satellite link architecture
Orbital altitude	500 to 1,200 km LEO; altitude selection trades drag, radiation lifetime and link budget
Satellite lifetime	5 to 7 years at typical LEO altitudes; replenishment cadence must be planned from programme outset
User terminal latency	Under 40 ms round-trip at 600 km altitude, a genuine advantage over GEO
Ground gateways	2 to 6 gateway stations depending on whether optical inter-satellite links are fitted
ITU filing lead time	Filing must precede first launch; coordination can take 3 to 7 years if contested
Programme timeline (constellation to operational)	5 to 8 years from contract to full regional service, including ITU coordination
National operations team	Typically 15 to 40 trained operators and engineers for a regional constellation at full handover
Indicative budget class	Several hundred million to low billions of dollars for constellation plus ground segment; IRIS2 at 10.6 bn EUR provides a continental-scale ceiling reference

Sources

One contract, one accountable engineer

Commissioned as one programme, not a stack of contracts: spacecraft, launch, ground segment, mission control, training and handover are priced together. Source-access terms and audit rights are agreed in writing before signature. Request a regional sizing assessment.