China's anti-Starlink push moves from theory to engineering

On 13 July 2026, the Russian-aligned Telegram channel Two Majors circulated a Chinese-language engineering paper arguing that SpaceX's Starlink low-Earth-orbit (LEO) broadband constellation could be disabled by a high-power microwave impulse fired from a ground-based phased array. The post, which framed the research as evidence that "space warfare is no longer the plot of science fiction films," is one of the clearest signals yet that Beijing's military-industrial complex is treating commercial mega-constellations as legitimate targets in any peer conflict, and is publishing the countermeasures.

The paper matters less for any single weapon concept than for what it reveals about Chinese strategic doctrine. China's published research establishment has spent two decades studying how the United States exploits space-based assets for intelligence, surveillance, reconnaissance, navigation, and (more recently) tactical communications to frontline units in Ukraine. A constellation that lets a soldier in a trench pull a 50-megabit-per-second downlink from orbit changes the math on every conventional engagement the People's Liberation Army (PLA) has planned since the late 1990s. Disabling it is no longer optional.

What the paper actually proposes

The text, summarised in the Two Majors channel and consistent with open Chinese-language journals indexed on CNKI and Wanfang, sketches a ground-based phased-array antenna firing a tightly focused burst of microwave energy into a LEO satellite as it passes overhead. The geometry is favourable: a Starlink shell sits at roughly 550 kilometres altitude, completing an orbit in about 95 minutes, which gives an observer at the equator roughly four passes a day per satellite, more at higher latitudes. The dwell time is short, but the energy budget needed to upset a satellite's receiver front-end or scramble its phased-array link back to ground is, according to the authors, plausibly achievable with commercially available components. The paper reportedly stops short of calling the system a weapon, framing it as "electronic countermeasure equipment" for testing, which is the standard Chinese formulation for dual-use space hardware.

Two engineering assumptions carry over from earlier Chinese work. First, that mega-constellations are soft targets at the radio-frequency layer because their antennas point downward at users, leaving their receivers exposed. Second, that the sheer size of a Starlink-style network, paradoxically, creates vulnerabilities: a constellation dense enough to provide global coverage is dense enough to be illuminated and tracked, and the software-defined radios on board can be spoofed into wasting their ephemeris budget chasing false handshakes.

Why a Russian milblogger is amplifying this

The framing in Two Majors is not idle curiosity. Russian forces in Ukraine have relied heavily on Starlink terminals since 2022 to backfill a domestic satcom capacity that has been degraded by sanctions on the Russian space industry and by the loss of several high-throughput geostationary birds to age and sabotage. Moscow's own counter-space programme, run out of the Aerospace Forces and Almaz-Antey, has historically prioritised anti-satellite (ASAT) missiles and on-orbit kinetic interceptors. A Chinese-style ground-based directed-energy solution would, in theory, give Russian units a non-kinetic option that does not produce the orbital debris that has already cost the ISS two avoidance manoeuvres in 2024 and 2025.

That shared interest explains why a Russian Telegram channel is the vector here, rather than an official MOD press release. Russian and Chinese commentary on counter-space has converged over the past eighteen months: both governments increasingly treat US commercial space infrastructure as a legitimate military target in any Taiwan or NATO scenario, and both are wary of testing that thesis with a kinetic strike because the debris field would compromise their own civil and military satellites.

The Western read, and why it is incomplete

Washington's response, in the handful of think-tank pieces and congressional hearings that have touched on the subject, is to point at the technical obstacles. High-power microwave weapons lose coherence over hundreds of kilometres of atmosphere; a satellite's casing provides roughly 15 dB of shielding against a ground burst; and Starlink satellites are designed to be replaced, with SpaceX currently launching more than 200 per month from Starbase and Cape Canaveral combined. None of that is wrong, but it is also the same dismissive tone Western analysts took toward Chinese hypersonic glide vehicles in 2014, before two test surprises in 2021 forced a rethink. The pattern is familiar: the dominant framework treats any adversary counter-space concept as either a propaganda bid or a primitive cousin of a US system, and is then embarrassed when the engineering turns out to work.

The structural point is that the US-led order in orbit rests on a small number of commercial choke points. Starlink, the OneWeb second-generation constellation, and the eventual Amazon Project Kuiper are private-sector assets that the US Department of Defense has bought into as core battlefield infrastructure through contracts with the Space Force's Commercial Augmentation Space Reserve. If a peer competitor can field a relatively cheap ground-based system that denies those constellations selectively, in a defined theatre, for hours at a time, the assumption that "space is a force multiplier we own" starts to wobble. That is the deeper argument the Chinese paper is making, regardless of whether the specific microwave concept ever flies.

What to watch

Three signals will tell whether this stays in the journals or becomes a programme. First, recruitment notices: if the China Electronics Technology Group Corporation (CETC) or the China Academy of Engineering Physics posts vacancies for high-power microwave engineers at the master's-and-above level, a paper programme is staffing up. Second, export filings: high-power vacuum tubes, gyrotron components, and large-format gallium-nitride radar wafers are dual-use items tracked under the Wassenaar Arrangement; sudden Chinese demand spikes show up in customs data within months. Third, exercise behaviour: PLA ground-force units have, since 2023, begun practicing electronic-warfare missions against simulated LEO downlink geometry. If those drills expand from jamming handhelds to firing high-power bursts at surrogate dishes, the engineering will have caught up with the doctrine.

What remains genuinely uncertain is whether China will deploy such a system openly or keep it as a strategic reserve. The first option invites US counter-proliferation pressure and a hardening of SpaceX hardware; the second preserves ambiguity, which is the currency Beijing has preferred to spend since the 1996 anti-satellite test fallout. The Chinese-language engineering literature tends toward ambiguity by design: every "electronic countermeasure" paper carries a disclaimer that the research is for the protection of Chinese space assets. Western readers who treat that disclaimer as sincere are repeating the mistake the Taiwan air-defence community made in the late 1990s, when it assumed that Chinese surface-to-air missile upgrades were defensive until they were not.

How Monexus framed this: the wire so far has run the Chinese paper as a curiosity item. We read it instead as a doctrinal marker, and paired the engineering specifics in the Russian milblogger summary with the broader PLA counter-space trajectory. The structural argument, that commercial space infrastructure has become a legitimate military target in any China-US scenario, is the thread that ties the piece together.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://t.me/two_majors
• https://en.wikipedia.org/wiki/SpaceX_Starlink
• https://en.wikipedia.org/wiki/Counter-space_weapon
• https://en.wikipedia.org/wiki/China_Electronics_Technology_Group
• https://en.wikipedia.org/wiki/Commercial_Augmentation_Space_Reserve