Satellite megaconstellations: the race to own low Earth orbit

What it is

The satellites-and-constellations beat tracks the race to build low-Earth-orbit (LEO) broadband networks, the global contest over spectrum and orbital slots, and the geopolitics of sovereign connectivity. A LEO satellite orbiting at 480 to 1,200 km delivers round-trip latency of 25 to 60 milliseconds, versus 600+ milliseconds at geostationary altitude. That physics gap made mobile, maritime, and rural broadband viable at mass-market scale for the first time. Four operators now hold the field: SpaceX's Starlink (US), Amazon Leo (US), China's state-linked Guowang and Qianfan programs, and the France-UK Eutelsat/OneWeb group.

History

The first serious megaconstellation proposal came from OneWeb founder Greg Wyler in 2012. SpaceX filed for its first Starlink satellites in 2016; two test vehicles launched in February 2018. Amazon filed US FCC applications in 2019. The ITU's "use it or lose it" deadline, first satellite within seven years and full constellation within 14, framed the competitive clock.

Three events in 2020 set the current race in motion. OneWeb filed for US Chapter 11 bankruptcy in March after SoftBank withdrew funding; the UK government and India's Bharti Global rescued it in July for US$500 million each; China in September filed ITU applications for 12,992 Guowang satellites and designated satellite internet national infrastructure. The US FCC authorized Amazon Kuiper's 3,236-satellite constellation that same July.

Russia's February 2022 invasion of Ukraine demonstrated Starlink's strategic weight and severed OneWeb from Soyuz rockets. Eutelsat merged with OneWeb in September 2023 (US$3.4 billion all-stock); Qianfan opened deployment in August 2024, Guowang in December 2024.

Current state

As of early July 2026, Starlink holds roughly 10,000 active satellites, with V2 Mini batches continuing to fly from Vandenberg (see فالكون 9 يحمل ستارلينك 17-45 على صاروخ في رحلته الخامسة والعشرين). It serves more than 9 million customers in about 155 countries, received US FCC authorization for an additional 7,500 satellites in January 2026, and has expanded its direct-to-cell service to roughly 25 million mobile beta users. Amazon Leo stands at 375+ satellites in orbit; the FCC waived the company's July 2026 half-constellation deadline but demoted spectral priority for satellites placed after the milestone. Qianfan passed 200 satellites in mid-2026 after a thruster-failure pause; Guowang batch launches have put roughly 190 satellites in orbit. OneWeb/Eutelsat operates 648 first-generation LEO satellites at 1,200 km alongside a legacy geostationary fleet.

Spectrum is the core constraint. All four operators hold ITU filings for overlapping Ku, Ka, and V frequency bands; China filed in 2020 for additional constellations totaling as many as 200,000 satellites. The ITU's updated milestone rule requires 10% deployment within two years and 50% within five, raising the cost of speculative filings.

Relationships

The four roster subjects interact on three axes. First, spectrum priority: ITU coordination grants seniority to whoever deploys first in any given band, making launch cadence a regulatory weapon. Second, launch supply: Amazon Leo's ramp depends on Ariane 6 cadence and US Vulcan capacity; Guowang and Qianfan together are projected to consume 70 or more of China's 2026 orbital launches. Third, customers: government and enterprise buyers across the Global South weigh Starlink's US-sovereign risk against Qianfan's Chinese-sovereign risk and against IRIS2, the EU sovereign constellation backed by Eutelsat.

OneWeb/Eutelsat bridges the commercial and sovereign tracks, contributing capacity to IRIS2 while competing directly with Starlink for maritime and enterprise customers. Orbital congestion is the shared constraint: ESA documents 28,000 tracked objects in LEO and collision-avoidance manoeuvres rising roughly 20% year-on-year, with megaconstellations as the primary driver (see وكالة الفضاء الأوروبية: خطر الاصطدام في المدار المنخفض ارتفع ~20% عام 2026 مع تجاوز الأجسام المرصودة 28,000).

What to watch

• Amazon Leo's beta service (targeted late 2026) in five countries: whether terminal hardware and latency match Starlink determines the consumer competitive dynamic.
• Guowang's ITU milestone: China SatNet must orbit roughly 6,500 satellites by 2032, requiring hundreds of launches per year from 2028, a cadence not yet demonstrated.
• Eutelsat's financial restructuring: a €960M March 2026 equity raise is the first of three transactions needed to fund OneWeb Gen-2 (440 satellites from late 2026) and the group's IRIS2 contribution (see Eutelsat تجمع أموالاً لتمويل الجيل الثاني من OneWeb وكوكبة IRIS² السيادية الأوروبية).
• The 2027 ITU World Radiocommunication Conference: competing US and Chinese positions on Ku/Ka/V band allocations will set every constellation's growth ceiling.
• Direct-to-cell expansion: Starlink's live T-Mobile partnership and Amazon Leo's carrier talks signal a next-phase market of hundreds of millions of users by 2028.