Cool War in LEO

The scenario

Two great powers can hurt each other in orbit every day and still keep the war cool. A Taiwan crisis is the most likely place that dynamic actually plays out: both sides have reasons to interfere with the other's satellites, and both have reasons to keep the interference deniable enough that no single incident forces a public response. The cool war isn't a diplomatic choice — it's what the capability picture in 2026 permits, and what neither side can easily escape.

Three lines of recent open-source writing converge on this thesis without yet using a single common label. Modern War Institute's 2025 essay Space as a Gray Zone argues that the defining traits of orbital warfare are now attribution opacity and dual-use ambiguity, and that nonkinetic effects — jamming, dazzling, cyber — leave no trace but achieve the same goal, while civilian and commercial cover lets state actors maneuver across the orbital chessboard unnoticed and unchallenged.¹ The companion MWI piece Red Lines in Orbit goes further: A disabled satellite can be framed as a technical failure or an act of war. That ambiguity is what makes space so strategically volatile.² The Air University China Aerospace Studies Institute's May 2025 brief on deterring PRC use of force in space treats reversible attacks — GPS jamming, optical dazzling — as the hardest deterrence problem because their low cost and proportionate optics collapse the deterrence equation.⁷

The frame is not new in family. James Finch's 2015 Joint Force Quarterly piece, Bringing Space Crisis Stability Down to Earth, was already flagging it: many space capabilities can be degraded through electronic means, enabling the use of weapons systems such as jammers that an adversary might perceive as less escalatory.⁵ A decade later, that argument has been operationalized in PLA doctrine, Russian electronic warfare against Starlink in Ukraine, and US Space Force “competitive endurance” theory. What has changed is not the analytic insight; what has changed is that the capability picture now makes the cool war the natural strategy on both sides at once.

The cool-war frame is not the only plausible reading of a Taiwan-adjacent crisis in space. The opposing scenario is the kinetic opening — a fast, large-scale PLA counterspace campaign aimed at blinding US ISR, PNT, and comms in the first hours of any move on the island. Wild Blue Yonder's January 2026 piece Phase One: Blinding the Eagle is the most developed open-source articulation of that alternative, written from a Chinese-planner perspective and laying out a 24-hour opening counterspace campaign before any kinetic action against Taiwan.⁴ The cool-war frame in this memo treats that scenario as the foil — the case where the deniability equilibrium has already collapsed, or where Beijing has decided to skip it entirely. The capability picture supports both readings; which one obtains depends on Beijing's read of US resolve and on whether the gray-zone tools have been exhausted before the crisis cuts in.

A working definition

What follows uses “cool war in LEO” as shorthand for a specific kind of contest: sustained, recognized by both sides, deliberately kept below the threshold of open war. Effects are reversible. Incidents are ambiguous. Signaling is visible to anyone watching, but the proof needed to compel a public accounting is not. Both parties prefer that arrangement to the costs of escalation. The label is not a term of art at West Point or Air University — the literature there uses “gray zone,” “sub-threshold competition,” “competitive endurance,” or “orbital security dilemma” for adjacent ideas.³⁶ None of them quite fits the dynamic this report examines, which is why we coin one.

Why now — structural drivers

Four shifts since the early 2020s have made low Earth orbit unusually well-suited to this kind of contest. The number of satellites has exploded. The line between civilian and military hardware has blurred to near-incoherence. Both sides have rebuilt their architectures around absorbing damage rather than preventing it. And the unwritten rules against orbital aggression have visibly weakened. The 2024–26 open-source record documents each.

Proliferation

The shift is from a small number of expensive, hard-to-replace satellites to vast networks of small, cheap, replaceable ones. The Space Development Agency (SDA), the Pentagon office building this proliferated layer, is targeting 154 operational satellites in its first tranche — 126 for data routing, 28 for missile-warning surveillance — at roughly monthly launch cadence, with initial warfighting capability scheduled for 2027.⁵⁰ Tranche 2 adds 200 more data satellites split across York Space Systems, Lockheed Martin, Northrop Grumman, and Rocket Lab, plus 54 surveillance satellites across L3Harris, Lockheed, and Sierra Space, at $3.5 billion total.⁵¹ Layered on top of the SDA effort, SpaceX's Starshield — the government-only variant of Starlink — is reportedly running an intelligence constellation of more than 150 satellites for the National Reconnaissance Office by 2025, with a separate Space Force communications program (MILNET) building on a previously undisclosed ~$1.8 billion Starshield contract.⁵⁴ Amazon's Project Kuiper began launching operationally in April 2025; the Pentagon's stated aim is to break SpaceX's near-monopoly on this layer.⁵⁵

China is following the same trajectory but lagging the schedule. Its flagship constellation, Guowang — the project of state-owned China Satellite Network, filed with the International Telecommunication Union (the ITU; covered in Part III) as 12,992 satellites — began launching in December 2024 and reached roughly 145 satellites in orbit by January 2026. That is well below the deployment pace the ITU rules require. A second Chinese mega-constellation, Qianfan (also called G60 or Thousand Sails), reached around 126 by April 2026 against a planned first phase of 1,296.¹⁰³¹⁰⁵ Across all Chinese filings, the paper total now approaches 200,000 satellites — most of which represent intent rather than hardware.²⁶ Both sides are building proliferated layers; the deployment curves are radically asymmetric.

Dual-use opacity

Inspector, servicer, and debris-removal satellites are functionally indistinguishable from weapons. The robotic arm that grapples a dead satellite to deorbit it can grapple a live one to disable it. The fuel-transfer mechanism that extends a friendly satellite's life can be aimed at an adversary's. China's Shijian-21 (SJ-21) demonstrated the dynamic most cleanly in January 2022, when it docked with and towed a defunct Chinese navigation satellite (Compass-G2, part of the BeiDou-2 system) roughly 3,000 km above the geostationary belt — a working capture-and-tow capability that China framed as routine debris management and Western analysts read as a potential coercive tool that can disable a satellite without leaving any tracked debris behind.³²³³ CASI's 2021 pre-event analysis of SJ-21 had already anticipated the demonstration: SJ-21 introduces a potential counterspace capability that would allow China to disable a satellite without generating dangerous space debris or the associated reputational cost.³² SJ-25, launched January 2025, conducted apparent on-orbit refueling with SJ-21 in 2025; commercial trackers (Slingshot, COMSPOC) called it the first practical GEO refueling demonstration, and US Space Force inspector assets repositioned to observe.³⁷³⁹ In LEO, USSF Vice Chief Gen. Guetlein in March 2025 publicly characterized observed PRC satellite behavior as dogfighting — three Shiyan-24C and two Shijian-6 satellites maneuvering in and out around each other, in synchronicity and in control.⁴¹ A Slingshot analysis published in April 2026 cataloged 75 unusual maneuvers by Chinese satellites over nearly a decade.⁴²

Resilience asymmetry

The US and its allies have built their architecture around routing around damage rather than preventing it. That changes Beijing's problem in a fundamental way: an attack that disables a few satellites does not disable the network. To matter, the pressure has to be sustained — which pushes Beijing toward tempo rather than magnitude. RAND's 2024 work on strategic stability captures the dynamic directly: communications channels with China are likely unavailable in a crisis, and decision cycles will be compressed with little warning.¹⁵¹⁶ The companion finding, in RAND's July 2025 commentary Why the United States Should Not Fear a Space Pearl Harbor, is that a single knockout blow in space is unlikely from either side — resilience and terrestrial backups absorb the effect of any one shot.¹⁷ The capability picture pushes both sides toward sustained pressure, not decisive single events.

Eroding norms

Three direct-ascent anti-satellite (ASAT) tests over fourteen years — China's destruction of its own FY-1C weather satellite in 2007, the United States' interception of the failing USA-193 in 2008, and Russia's destruction of Cosmos 1408 in 2021 — have steadily weakened the assumption that low Earth orbit is a sanctuary. Disputes over the ITU's 2019 deployment-milestone rules and increasingly close-approach maneuvers between rival satellites (rendezvous and proximity operations, or RPO) have eroded the day-to-day rules of the road on a separate track. In April 2022 the United States announced a unilateral commitment not to test destructive direct-ascent ASAT systems again; later that year UN General Assembly resolution 77/41 endorsed the same moratorium internationally, passing 154 in favor, 8 against, 10 abstentions. The eight “no” votes were Belarus, Bolivia, China, Cuba, Iran, Nicaragua, Russia, and Syria.⁸¹⁸⁵ The norm now exists on paper, but the principal actors whose behavior the cool-war thesis depends on are not in it. The Secure World Foundation counts roughly 38 states making the unilateral commitment by late 2024.⁸⁷ SWF's eighth annual Global Counterspace Capabilities report, released April 2025, totals the cumulative ASAT-test debris ledger across the US, Russia, China, and India: 6,851 cataloged trackable pieces, of which 2,920 remained in orbit as of February 2025.¹¹¹²¹³ Aerospace Corporation's Space Agenda 2025 describes the same period as one in which Russian and Chinese counterspace programs increasingly mirror each other: advanced on-orbit capabilities which could serve as inspection and repair satellites or co-orbital weapons alongside dedicated electronic-warfare and jamming systems.¹⁴

The legitimacy layer (ITU)

The ITU cannot physically prevent harassment. What it does is supply a political story — one that raises the cost of misbehavior for legitimate operators and, conversely, provides a cover narrative for a defector. Understanding how that mechanism actually works is load-bearing for everything that follows.

WRC-19 Resolution 35 — what it actually says

At the 2019 World Radiocommunication Conference (WRC-19), the ITU passed Resolution 35, which requires operators of non-geostationary satellite systems — meaning almost every new mega-constellation — to deploy 10% of their authorized satellites within two years of activation, 50% within five years, and 100% within seven.⁹⁸¹⁰⁰ Missing those milestones does not trigger fines or expulsion. The mechanism is administrative: the ITU Radiocommunication Bureau (the BR) reduces the number of satellites recorded in the Master International Frequency Register (the MIFR) to the count actually deployed at the milestone date, capping how many satellites can claim ITU-protected priority.⁹⁹ The policy reasoning, as then-BR Director François Rancy put it ahead of WRC-19, is to stop operators from warehousing spectrum: an ITU filing costs roughly 20,000 Swiss francs, and compared to the cost of the satellite and the launch, the regulatory fee is close to zero, and in return you get access to valuable spectrum resources pretty much forever.¹⁰¹

The 2023 conference (WRC-23, Dubai, November–December) tightened the regime further.¹⁰⁹¹¹⁰ Non-geostationary satellites must now stay within 70 km above or below their registered orbital position during deployment and 30 km thereafter — a structural fix against the practice of filing for one orbit and quietly drifting to another.¹⁰⁷ Resolution 35 itself was revised, and the ITU's research arm (ITU-R) was directed to study what happens after the seven-year deadline — meaning the enforcement mechanics past that mark are still being drafted.⁹⁹ A separate ongoing dispute over interference limits between non-geostationary and geostationary operators (the equivalent power flux density rules in the ITU Radio Regulations, Article 22) was extended toward the 2027 conference with no immediate regulatory consequence — a working compromise between the new LEO entrants and the geostationary incumbents.¹⁰⁸

The single most important fact about Resolution 35 is also the easiest to miss: there is no documented public case, since 2019, of the BR formally applying the milestone-reduction mechanic against a major operator. The earliest hard 10% deadlines for most post-2020 mega-constellation filings only fall due around 2027–2029. The teeth of the regime have not yet been tested.

PRC filings — what is actually on file, and what is in orbit

What “China defects from ITU” actually means

CSIS analysts Young and Thadani frame the stakes precisely: once operators have occupied a large amount of orbital and spectrum resources, other countries need to avoid the frequency bands and orbital positions that have been applied for to avoid mutual interference.¹¹² CSIS's broader analysis of the spectrum-allocation contest with China extends the point: the ITU is a forum where China is structurally advantaged — through coordinated state delegations, bloc voting, and filings backed by state-owned enterprises.¹¹¹ Lawfare's legal-statecraft work argues the broader pattern is “semantic lawfare”: embedding preferred legal constructs in international fora, with no organized US counter-strategy.¹¹³ A 2022 War on the Rocks essay on amending the 1967 Outer Space Treaty captures the structural problem one level up: key treaty terms are undefined (“harmful interference,” “national appropriation”), there is no agreed enforcement venue, and there are no rules of the road that bind states the way they bind private operators.¹¹⁴

Applied to ITU specifically, defection means three concrete behaviors, none of which require leaving the institution. First: ignore MIFR reductions that disadvantage them. Second: file aggressively, occupying orbital and spectrum positions regardless of any real intent to deploy. Third: lean on national licensing through China's Ministry of Industry and Information Technology (MIIT) to operate domestically and over Belt and Road Initiative partner states even where ITU coordination is disputed. The ITU's tools — coordination, registration, milestone-based MIFR pruning — discipline polite operators. They do not punish a state actor willing to wear the cost of being uncoordinated.

The PRC capability picture

The most consequential capability change in counterspace this decade has happened on the Chinese side. In April 2024 the People's Liberation Army was reorganized to put space, cyber, and information warfare under direct top-level command. The Guowang and Qianfan mega-constellations are climbing slowly toward operational scale. China's Shijian and Shiyan inspector satellites have built a working library of close-approach and capture-and-tow techniques. The BeiDou navigation network supports a maturing offensive jamming and spoofing toolkit. And a coherent war concept — system destruction warfare — assigns space a paralyzing role in any future fight.

Architecture and actors — what the April 2024 reorg actually did

On 19 April 2024 Beijing dissolved the People's Liberation Army (PLA) Strategic Support Force, the umbrella body that had handled space, cyber, and electronic warfare since 2015. In its place came three separate arms reporting directly to the Central Military Commission, the body that runs the entire Chinese military: the PLA Aerospace Force (天军), the PLA Cyberspace Force, and a newly created PLA Information Support Force. The 2024 edition of the Pentagon's annual China Military Power Report is the first to formally describe the new architecture and offers two readings of the change — the dissolution raised suspicions about operational effectiveness and leadership issues among SSF top leadership, including likely corruption, while another reading holds that the new structure showed the CMC's interest in directly controlling the most critical strategic capabilities.²¹ The 2025 edition extends the discussion in the same framing.²² CSIS reads the move as Xi pulling space, cyber, and information functions one bureaucratic layer closer to himself, not as evidence of dysfunction.²³ The International Institute for Strategic Studies (IISS) calls it the most consequential PLA structural change since the 2015–16 reforms; a June 2024 War on the Rocks piece on the new Information Support Force adds operational texture.²⁴²⁵ The US–China Economic and Security Review Commission (USCC), in its 2025 chapter The Final Frontier, frames Guowang and Qianfan as an explicit strategic effort to crowd out U.S. companies and dominate space in the near term.³¹ The operational implication is the one that matters here: the friction that used to sit between China's space combatants and its conventional theater commands has been reduced. Capabilities on paper haven't changed; capabilities on call have.

Inspector satellites — the working capability library

By 2025 China had assembled a working library of close-approach and rendezvous behaviors at every orbit that matters. The China Aerospace Studies Institute (CASI), the US Air Force's open-source PLA-watching shop, published a March 2022 catalog of the Shijian (SJ) and Shiyan (SY) satellite series that remains the cleanest open-source mapping of which tail numbers fall into which mission family — close-approach, signals intelligence, or technology demonstration.³⁴ Brian Chow's Strategic Studies Quarterly piece Stalkers in Space remains the canonical analytic frame for treating co-orbital close approaches as a coercive-capability problem rather than a technical one.⁹ SJ-21 demonstrated geostationary capture-and-tow in January 2022.³³ SJ-17, launched in November 2016, executed multiple close approaches (within 55 km) to other Chinese geostationary satellites — CSIS treats it as the foundational tradecraft that flowed into SJ-21 and SJ-25.³⁵ SJ-25, launched in January 2025, conducted apparent geostationary refueling operations with SJ-21 through the summer of 2025; in the open-source consensus, this is the first practical orbital refueling demonstration anywhere.³⁶³⁸ The TJS series — TJS-3 in 2018 released a coordinated subsatellite widely cited as China's first dual-spacecraft proximity-and-spoofing demonstration; TJS-15 launched in 2025 — operates from geostationary orbit with classified payloads.⁴⁰ In low Earth orbit, a 2024–25 “dogfighting” demonstration involving five Chinese satellites was officially confirmed by the US Space Force in March 2025.⁴¹ Across nearly a decade, Slingshot Aerospace has cataloged 75 unusual maneuvers by Chinese satellites — the open-source picture is no longer of one-off demonstrations but of an operational repertoire.⁴²

Electronic warfare and cyber

Kristin Burke's December 2023 CASI brief is the best single open-source treatment of who controls which counterspace effect inside the PLA. Burke distinguishes between jammers a regional theater commander can task on his own authority (operated by non–Strategic Support Force services with central approval) and beyond-theater radio-frequency and cyber capabilities reserved for higher commands. PLA training-academy texts, she notes, describe attacking adversary satellites by intruding on their telemetry, tracking, and control (TT&C) ground stations as a primary technique.⁸ Burke concludes:

PLA academy texts describe degrading satellite services by way of intrusion into an adversary's telemetry, tracking, and control (TT&C) ground stations as a central counterspace technique.CASI — PLA Counterspace Command and Control (Burke, December 2023)

The implication is important: open sources treat ground-segment cyber as a doctrinally claimed capability whose actual demonstrated reach must be inferred from PLA academy texts and exercise reporting rather than from disclosed incidents. CASI's May 2024 BeiDou brief frames China's positioning, navigation, and timing (PNT) network as a warfighting backbone — with an indispensable role in multidomain precision warfare — and surfaces a growing PLA literature on offensive PNT, including jamming and spoofing aimed at GPS users.⁴³ CSIS reporting on Chinese electronic-warfare and communications buildup near the South China Sea has documented mobile GPS jammers on Mischief Reef since 2018.⁴⁴

Doctrine — system destruction warfare and its successors

The clearest Western articulation of China's underlying war concept remains Jeffrey Engstrom's 2018 RAND monograph Systems Confrontation and System Destruction Warfare. The PLA's theory of victory, in Engstrom's translation, is to paralyze and even destroy the critical functions of an enemy's operational system so that the enemy loses the will and ability to resist.⁴⁶ The concept is explicitly multi-domain — space, cyber, electromagnetic, and psychological warfare working together against the same target — and explicitly aimed at the enemy's system as a whole rather than at any one component.⁴⁶ A companion RAND volume by Pollpeter, Chase, and Heginbotham on the now-disbanded Strategic Support Force argues that the SSF was the organizational expression of the system-destruction concept — the body that coordinated space, cyber, and electronic warfare to paralyze the enemy's operational system-of-systems.⁴⁷ A 2020 RAND update introduces “multi-domain precision warfare” as the doctrine's post-2020 evolution.⁴⁸

A newer concept, intelligentized warfare (智能化战争), layers onto this. War on the Rocks summarizes its four characteristics as faster information processing, faster decision-making, swarming tactics, and cognitive effects — with AI-based space confrontation explicitly named as an operational concept.⁴⁹ Whether intelligentized warfare is mature doctrine or aspirational rhetoric is genuinely contested in the open source. The older system-destruction concept is not.

The US / allied capability picture

The US side has reorganized faster than the Chinese side at the doctrinal level but is still mid-transition at the architectural one. The decisive shift — outsourcing resilience to commercial mega-constellations — is also the decisive vulnerability, and the contested procurement fights of 2025 reflect that unresolved trade-off.

Architecture and actors

The US Space Force (USSF), US Space Command (USSPACECOM), the National Reconnaissance Office (NRO), the Space Development Agency (SDA), and a growing roster of commercial partners now make up a hybrid public-private architecture. The Space Force's chief of operations, Gen. Chance Saltzman, has built a “Competitive Endurance” theory of victory around three tenets: avoid operational surprise, deny first-mover advantage, and conduct counterspace operations responsibly — meaning, no long-lived debris.⁶⁸⁶⁹ The Space Force Doctrine Document 1 (SFDD-1), signed 3 April 2025, is the first major rewrite of the 2020 Spacepower capstone and elevates space control as a core function: the activities required to contest and control the space domain, with space superiority as the desired outcome.⁶⁶⁶⁷ Service Doctrine Publication 3-0 (SDP 3-0, July 2023) is the connective tissue between the capstone doctrine and the operational level.⁹⁶ Under the four-star command of Gen. Stephen Whiting, USSPACECOM has framed orbit as a contested operating environment with a common playbook for warfighting advantage.⁷⁰

The proliferated layer

The SDA's September 2025 fact sheet for the first tranche of its Proliferated Warfighter Space Architecture locks the numbers: 154 operational satellites (126 for data routing, 28 for missile-warning surveillance) launched in roughly ten batches at monthly cadence, with initial warfighting capability scheduled for 2027.⁵⁰ The first launch went up from Vandenberg Space Force Base in September 2025; the second, with 21 satellites, on 15 October 2025.⁵⁰ The second tranche spreads the work across more contractors: 200 data satellites split between York Space Systems (62 of the lighter “Alpha” variant, ~$615 million), Lockheed Martin (36 of the heavier “Beta” variant, ~$816 million), Northrop Grumman (36 Beta, ~$733 million), and Rocket Lab; and 54 surveillance satellites across L3Harris, Lockheed, and Sierra Space at $3.5 billion in total contract value, with launches no later than April 2027.⁵¹

Layered on top of the government architecture is the commercial layer. SpaceX's Starshield is the government-only variant of Starlink; the publicly reported NRO contract underwrites a dedicated proliferated LEO intelligence constellation that exceeded 150 satellites by 2025.⁵⁴ The Space Force's Proliferated Low Earth Orbit Satellite-Based Services contract has a $13 billion ceiling over ten years, with an early $70 million task order to SpaceX containing unique terms and conditions not present in earlier commercial contracts — a direct policy response to the Ukraine experience that SpaceNews and Air & Space Forces Magazine both flagged in real time.⁵³⁵⁷⁵⁸ MILNET, reported in 2025, is a separate Space Force–SpaceX communications effort that sits alongside the SDA data layer or partly substitutes for it.⁵⁴ The most analytically interesting procurement fight of 2025 was the running question of whether out-tranche SDA data-layer purchases would simply be replaced by SpaceX commercial services — Breaking Defense reported in March 2025 that the Air Force was actively considering exactly that substitution.⁵² Amazon's Project Kuiper began operational launches in April 2025; the Defense Innovation Unit's Hybrid Space Architecture vendor pool (with Eutelsat and OneWeb participating) is the structural answer to the SpaceX-monopoly problem.⁵⁵⁵⁶ Coalition context matters here too: an IISS January 2025 paper on European space capabilities documents the post-Ukraine recalibration of European requirements, and the European Union has committed an additional €4.2 billion ($4.74 billion) for 2026–2030, with sentinel-satellite launches scheduled for 2027.¹⁹

Tactically Responsive Space

Tactically Responsive Space — the ability to put a useful satellite over a target on hours' notice rather than months — has its proof point in an exercise called Victus Nox. From the moment the Space Force gave the “go” call, Firefly Aerospace and Millennium Space Systems encapsulated, mated, transported, and launched the satellite from Vandenberg in 27 hours (24 hours of work plus a launch-window wait); the spacecraft was operationally ready 37 hours after launch.⁵⁹ Its successor, Victus Haze, slipped into 2026 after a Firefly Alpha rocket anomaly. Haze is the doctrinal step beyond Nox: two payloads (one from Rocket Lab, one from True Anomaly, on separate launchers) executing on-orbit maneuver and surveillance against each other to demonstrate, under operationally realistic conditions, U.S. Space Force's ability to respond to irresponsible behavior on orbit.⁶⁰ Doctrinally, Tactically Responsive Space is now treated as a behavior modifier — the option to put eyes on a target on hours' notice — rather than a primary warfighting tool.

Space situational awareness

At geostationary altitude, the six-satellite Geosynchronous Space Situational Awareness Program (GSSAP) provides the foundation for tracking adversary behavior.⁶¹ Silent Barker (mission designation NROL-107, launched 10 September 2023 on an Atlas V rocket, with full operational capability targeted for 2026) is described in open reporting as an exponential leap in America's capability to keep tabs on potentially threatening Russian and Chinese satellites, carrying multiple sensor payloads in geostationary orbit.⁶²⁶³ The Deep Space Advanced Radar Capability (DARC) is a three-site radar built jointly under AUKUS — the Australia–UK–US security partnership — with sites at Exmouth in Western Australia, in the United Kingdom, and on the US continent. Site 1 broke ground in October 2023, completed construction in December 2024, ran a first seven-antenna tracking demonstration in August 2025, and targets full operational capability at site 1 in 2027, with the full system in 2032.⁶⁴ The hardest gap to close is fast attribution in low Earth orbit; that is the job Joint Commercial Operations (JCO) is trying to do, by aggregating data from LeoLabs (six radar sites tracking objects as small as 2 cm in LEO), Slingshot Aerospace (AI-driven tagging of maneuvering and cloaking objects), ExoAnalytic Solutions (optical and passive radio-frequency tracking of geostationary objects), and COMSPOC, with a late-2025 reorganization aimed at pushing commercial data into operational tasking.⁶⁵

Defensive and offensive space EW

The Counter Communications System Block 10.2 (CCS Block 10.2) reached initial operating capability on 9 March 2020. It is a transportable, ground-based radio-frequency jammer designed to deny an adversary's satellite-communications downlink, built by L3Harris.⁷¹⁷² Meadowlands, its lighter and more mobile successor, was approved for Space Force training use on 2 May 2025 and offers multi-frequency S- and X-band jamming.⁷³⁷⁴ Bounty Hunter is a satellite-communications signal-monitoring and geolocation system being transitioned to Space Force Delta 3 — primarily defensive in posture, complementary to the offensive jammers.⁷⁴ Together, CCS, Meadowlands, and Bounty Hunter make up the publicly acknowledged US offensive-and-defensive space electronic-warfare toolkit.

Capability flashpoints

Five places where the two pictures actually meet. None of these is a single discrete “contest;” each is a layered competition where capability investments on one side trigger countermeasures on the other.

1. Communications degradation

This is the headline flashpoint. China's jamming and spoofing capability against satellite communications — now distributed across the new Aerospace, Cyberspace, and Information Support Forces after the 2024 reorganization — is real and demonstrated. Russian operational use against Starlink in Ukraine has shown that proliferated commercial LEO communications services can be effectively jammed at the user-terminal level, which is a different and harder problem than attacking the satellite itself.¹¹⁵ Western routing around damage through proliferated networks is the structural answer; the unresolved question is whether terminal-level resilience can keep up with the iterative pace of electronic-warfare improvements on the other side. SpaceX's March 2022 software-patch response — US Department of Defense electronic-warfare director Dave Tremper called it eyewatering — was an early data point, but by 2024 Russian success against Ukrainian Starlink was reportedly recurring with the so-called “Kalinka” jamming system.¹¹⁵¹¹⁶

2. Navigation and timing

The US is structurally exposed to spoofing of China's BeiDou navigation system and to ground-based jamming, because precision-guided weapons, civil aviation, and most modern logistics depend on global navigation satellite signals. CASI's May 2024 brief frames BeiDou as a warfighting backbone; a May 2025 War on the Rocks piece, A Signal Point of Failure, argues that the integration of BeiDou into US navigation systems is itself a vulnerability.⁴³⁴⁵ The US response — multi-system receivers that can fall back from one constellation to another, military-only encrypted GPS signals (M-code), and terrestrial backups built around enhanced long-range navigation (eLORAN) — is incomplete and fielded unevenly across the services.

3. Close approaches in orbit

China's inspector-satellite cadence is rising in both low Earth orbit and geostationary orbit. The US response — GSSAP at geostationary altitude, Silent Barker, the Joint Commercial Operations cell, and Tactically Responsive Space as a behavior modifier — is real but mid-build. The capability gap that matters most is fast attribution in low Earth orbit: Silent Barker and DARC are aimed at geostationary orbit, but the harder problem is identifying who did what at LEO altitude, where orbital periods are roughly 90 minutes and a satellite can be over a target and gone before commercial sensors finish tagging it. That is the problem JCO and Tactically Responsive Space are collectively trying to solve.⁶⁵

4. Cyber

China's cyber capability against US space ground segments is, in open source, more a doctrinal claim than a demonstrated record. The flip side — the cyber posture of US commercial space operators — is the soft underbelly. The Starshield contract's “unique terms and conditions” are the Pentagon's emerging policy answer; Kuiper and OneWeb diversification are the structural one.⁵³⁵⁵

5. Targeting and surveillance

Both sides use space to find each other. The real contest is not for orbits but for the kill chain — the closed loop from sensor to shooter to target effect — and which side can keep that loop closed under contested conditions. RAND testimony by Brian Alkire to the US–China Economic and Security Review Commission in April 2025 captures the headline trend: China's intelligence, surveillance, and reconnaissance (ISR) satellite count has grown roughly sixfold over eight years, including roughly seventeenfold growth in commercial ISR.¹⁸ The targeting-support contest is the one most directly continuous with conventional military operations — and the one most likely to produce the moment where the cool war becomes a shooting war.

Tripwires — capability-grounded

Three tripwires dominate the doctrinal and analytic literature: a mass-debris event, an effect on a strategic system, and high-confidence attribution. None of them is an abstract red line. Each is grounded in specific capability decisions, and each can be engineered closer to or further from the threshold.

Tripwire 1 — Mass debris

The November 2021 Cosmos 1408 incident remains the cleanest example. Russia's direct-ascent anti-satellite test that day produced, in the public assessment of US Space Command, more than 1,500 pieces of trackable orbital debris and hundreds of thousands of pieces of smaller orbital debris that now threaten the interests of all nations.⁷⁵ The crews of the International Space Station's Crew-2 mission and the docked Soyuz spacecraft were ordered to shelter in their capsules through two passes of the cloud.⁷⁶ NASA's Orbital Debris Program Office (ODPO) tracked the rapid decay of the Cosmos 1408 fragment cohort: roughly 450 cataloged fragments remained one year after breakup, about 90 by two years, and the cataloged-fragment curve effectively ended on 3 February 2024. The cloud decayed faster than China's 2007 FY-1C cohort because Cosmos 1408 broke up at lower altitude (~480 km versus ~865 km).⁷⁷

FY-1C is the worse case. NASA describes the 2007 Chinese test against the FY-1C weather satellite as the most severe artificial debris cloud in Earth orbit since the beginning of space exploration, with more than 2,000 debris on the order of 10 cm or greater in size identified by the U.S. Space Surveillance Network.⁷⁸ NASA ODPO's 2015–2025 comparison of debris density below 1,200 km shows the FY-1C cloud (peaking at ~750 and ~850 km) mostly unchanged — that population is going to be there for a long time. The European Space Agency's 2025 Space Environment Report frames the same picture as system-wide pressure: roughly 40,000 tracked objects in orbit (about 11,000 active payloads), an estimated 1.2 million objects larger than 1 cm, more than 50,000 larger than 10 cm, and intact satellites or rocket bodies now reentering the atmosphere on average more than three times a day. In the most heavily populated altitude bands, the density of active satellites is now the same order of magnitude as the density of debris.⁷⁹⁸⁰ LeoLabs's 2022 LEO annual review and Darren McKnight's March 2024 debris-management work add the operational picture: by year-end 2022 the ratio of conjunctions driven by space-traffic management to those driven by debris management had inverted from 60/40 to 80/20, and the 800–900 km band remains the “bad neighborhood” that contains a third of the highest-risk close approaches.⁹⁰⁹¹ The point for the cool war is structural: kinetic anti-satellite use is bounded by the attacker's own future operating environment. China's 2007 demonstration produced a debris cloud that constrains Chinese operations today. Russia's 2021 test attracted near-universal condemnation and accelerated the multilateral norm-building track.

The 2022 US ASAT moratorium and UNGA Resolution 77/41 (adopted 7 December 2022, vote 154–8–10) are the multilateral response.⁸¹⁸² Arms Control Association tracking documents the early-adopter sequence (US in April 2022, then Canada in May 2022, New Zealand in July 2022, Japan and Germany in September 2022, UK and South Korea in October 2022, with seven additional countries joining by November 2022).⁸³⁸⁴ Netherlands, Austria, and Italy added momentum in early 2023, becoming the eleventh through thirteenth signatories.⁸⁶ Approximately 38 states have made the unilateral commitment as of late 2024 per Secure World Foundation tracking.⁸⁷⁸⁸ The eight UNGA “no” votes in 2022 were Belarus, Bolivia, China, Cuba, Iran, Nicaragua, Russia, and Syria.⁸⁵ The norm exists, with measurable adherence; the principal counterspace actors are not in it.

Tripwire 2 — A nuclear-related satellite gets touched

This is the hardest tripwire to characterize cleanly — partly by design. The 2022 National Defense Strategy and Nuclear Posture Review (NPR) keep the deterrence calculus for non-nuclear strategic attacks (which would include attacks on satellites supporting nuclear command, control, and communications — NC3) deliberately ambiguous, while explicitly committing to deter nuclear employment of any scale directed against the U.S. homeland or the territory of Allies and partners, whether on the ground, in the air, at sea, or in space.⁹⁷ The 2022 NPR retains the position that the US will use nuclear weapons only in extreme circumstances to defend the vital interests of the United States or its Allies and partners, without specifying space-asset thresholds.⁹⁷ Joint Publication 3-14, the US doctrine document on space operations, establishes that attacks on US space systems may warrant a response in any domain, and is deliberately non-specific about cross-domain thresholds.

The capability angle is straightforward: damage to or interference with NC3 satellites, missile-warning satellites, or other strategically important space systems is shaped in significant part by which constellations happen to be flying in proximity to NC3-relevant payloads at any given moment. A close-approach maneuver too near a US missile-warning satellite reads as a tripwire event whether the maneuvering satellite intended harm or not. Saltzman's “avoid operational surprise” tenet of Competitive Endurance is the operational-doctrine bet that this tripwire holds — that the US sees coming what it most needs to deter.⁶⁹

Tripwire 3 — Confident attribution

The Aerospace Corporation's open-source assessment captures the attribution problem in two sentences. On reversibility: jamming is usually completely reversible because once a jammer is turned off, communications can return to normal. On attribution: attribution of jamming can be tough because the source can be small and highly mobile, and users operating on the wrong frequency or pointed at the wrong satellite can jam friendly communications.⁹² On cyber: accurate and timely attribution of a cyberattack can be difficult, if not impossible, because attackers can use a variety of methods to conceal their identity.⁹³ The Secure World Foundation captures the close-approach version of the same dilemma: RPO activities might involve getting up close to a satellite to inspect it, to listen to what it is listening to, to jam communications coming from it, to interfere with its optical sensors, or other purposes.⁸⁹ The activity is observable. The intent is not.

The open-source consensus on attribution timelines is rough: kinetic and debris events are attributable within hours to days through US Space Surveillance Network tracking; jamming takes hours to weeks and is often inconclusive; cyber is “weeks to never.” The capability investment that matters most here is unclassified attribution — the kind that can be cited in public statements without compromising sources. Better space situational awareness raises the political price of any aggressive action, and so changes the cool-war calculus from the supply side.

Past the threshold — three branches

The tripwires don't lead anywhere automatic. Each one collapses a piece of the equilibrium, but what happens next depends on which property broke and on how the moment is read in both capitals. Three branching paths are visible in the open-source record, and none of them are mutually exclusive.

Acknowledged contest in orbit. The most likely branch. Once attribution is on the public record — or a debris event makes denial impossible — both sides have to operate openly, which can paradoxically keep the fight bounded by the orbital domain. Jamming campaigns get acknowledged. Anti-satellite options get publicly threatened. Coalitions form around the moratorium framework. The cool war's deniability premise is gone, but the contest still has a ceiling. This is closer in pattern to the Cuban Missile Crisis than to general war: brinkmanship at high stakes, with diplomatic channels active and the option to step back preserved.

Spillover into a terrestrial fight. The middle branch. US space doctrine — including Joint Publication 3-14 and the 2022 National Defense Strategy — explicitly preserves the option of cross-domain response to attacks on US space systems.⁹⁷ If a tripwire event is read in either capital as part of a deliberate move toward a Taiwan-Strait crisis (or vice versa), the orbital contest collapses into the larger fight: cyber operations against terrestrial targets, conventional posture changes in the Indo-Pacific, possibly limited kinetic action in a maritime theater. The space contest becomes one front of a larger war, and the cool-war equilibrium ends the way most equilibria end — by no longer being the central question.

Mutual de-escalation back to the equilibrium. The quietest branch, and historically common. If the tripwire event is ambiguous enough that one side can step down without admitting it crossed a line, or if the visible consequences of further escalation persuade both capitals that the costs are not worth it, the cool war resumes — at higher operating tempo and with less mutual trust, but it resumes. The 1983 Able Archer scare is the canonical analogue: a moment of acute strategic alarm followed by a deliberate, mostly silent return to the prior posture. Its orbital equivalent might be a confirmed close-approach incident that goes uncondemned in public diplomacy because neither capital is willing to start the next move.

There is a fourth branch the open-source literature handles with particular caution: a genuine miscalculation involving NC3, or a debris event affecting nuclear-relevant forces, that pulls strategic-posture logic into the picture before either side has time to read the situation. That path is what strategic-stability researchers worry about most, and is also the path where careful analysis runs into responsible-disclosure limits. A future report in this series will pick it up.

The point of naming these branches is not to predict which one obtains. It is to make clear that the cool war's collapse is not a single event with a single sequel. Which path the scenario falls into is itself the question strategic planners on both sides are silently asking every day the equilibrium holds.

The Ukraine case study

Ukraine is the closest case study available for proliferated commercial mega-constellations under wartime conditions. Three lessons from the Russian–Starlink confrontation and the SpaceX–Crimea episode now structure US policy responses and Chinese strategic discomfort.

Russian electronic warfare against Starlink

Bart Hendrickx's two-part Space Review series remains the most thorough open-source order-of-battle for Russian counterspace electronic-warfare systems: Tobol (a hybrid anti-jamming and uplink-jamming system fielded around 2018), Tirada-2 (a mobile satellite-uplink jammer), Bylina (an automated electronic-warfare command-and-control system), Krasukha-4 (a ground-based radio-frequency jammer with overhead applications), and the older R-330Zh “Zhitel” (a cellular and GNSS jammer).¹²¹¹²² A 2024 SpaceNews piece consolidates leaked-document evidence that at least three of the seven known Tobol installations were tasked specifically against Starlink. Tobol was originally designed to protect Russian satellites from foreign jamming; it has been repurposed to attack foreign communications-satellite links and navigation signals.¹¹⁵ The “Kalinka” system, fielded by the Russian Center for Unmanned Systems and Technologies in late 2024, was marketed openly as a “Starlink killer” and reportedly produced recurring Ukrainian-side outages by mid-2024.¹¹⁵

The early-war lesson was about software speed. Theresa Hitchens's April 2022 reporting captured the reaction of the Pentagon's electronic-warfare director, Dave Tremper, to SpaceX's next-day software patch defeating an initial Russian jamming campaign: The way that Starlink was able to upgrade when a threat showed up, we need to be able to have that ability.¹¹⁶ By 2024 the “eyewatering” first-mover advantage was eroding. Russia iterated on uplink jamming and beam-interference techniques; Breaking Defense's December 2023 reporting captured the Russia–Ukraine race-to-out-innovate dynamic that has shaped the electronic-warfare front since.¹¹⁷

Commercial-operator wartime discretion — the Crimea episode

The canonical inflection point came at Sevastopol. Walter Isaacson's biography of Elon Musk revealed that, in September 2022, Musk personally denied a Ukrainian request to extend Starlink coverage so Ukraine's naval drones could complete a strike on Russian Black Sea Fleet vessels in Crimean waters. Musk's stated reasoning was that extending coverage would have rendered SpaceX explicitly complicit in a major act of war. Senator Jack Reed (D-RI), then chair of the Senate Armed Services Committee, opened an inquiry in September 2023.¹¹⁸ SpaceX's chief operating officer, Gwynne Shotwell, told Breaking Defense in February 2023 that SpaceX didn't intend that Starlink be ‘weaponized.’¹²⁰ The mirror-image problem — Russian forces obtaining Starlink terminals through black-market routes — required a separate Pentagon–SpaceX takedown effort that the Pentagon characterized as a cat and mouse game.¹¹⁹

Starshield as the policy answer

The “unique terms and conditions” language in the Starshield contract, and the Pentagon's push for Kuiper and OneWeb as alternative suppliers, are direct policy responses to the Crimea episode.⁵³⁵⁵ The structural shift is from a model where the operator decides to one where the customer decides: Starshield's integration of classified payloads, government-owned cryptographic keys, and contractually controlled coverage move the in-theater decision lever from the company to the contracting government. The policy is brand new, and untested in a Pacific scenario.

The Taiwan analog

RAND's 2024 report Chinese Military Views of Low Earth Orbit: Proliferation, Starlink, and Desired Countermeasures (Wang, Smith, Garafola) is the single most useful document for bridging the Ukraine experience to a Taiwan scenario. It documents PLA writings responding directly to the Starlink-in-Ukraine demonstration: Project SatNet (China's domestic “Starlink” equivalent) has been elevated to a national priority, and the broader phenomenon of proliferated LEO is described in PLA literature as directly undermining the PLA's preferred “informationized operations” playbook.¹²³ The key analytic finding for a Taiwan scenario: Taiwan itself has been reluctant to adopt SpaceX, partly because of Tesla's exposure to the Chinese market and partly because the Crimea precedent — the operator's option to remotely shut off coverage — is exactly the wrong lesson for an island that needs to assume coverage will be sustained.¹²³ CSIS Interpret: China's translated PLA essays on “Starlink militarization” are the primary-source Chinese voices on the same theme.¹²⁵ A 2024 SpaceNews report on a US wargame found that a PRC opening move combining electronic warfare with undersea-cable cuts could collapse Taiwan's connectivity absent a proliferated LEO substitute.¹²⁴

Implications for planning

If the cool war is the central case — which the capability picture, the doctrinal picture, and the Ukraine experience together suggest it is — six investments are decision-relevant. Two more look attractive on paper but don't change the equilibrium.

Investments that pay

Space situational awareness in low Earth orbit. Silent Barker and DARC are aimed at geostationary altitude; the gap that matters most is at LEO altitude, where attribution has to happen in minutes, not days. Joint Commercial Operations is the right structural answer; the open question is whether commercial fusion can actually deliver minutes-fast attribution under contested conditions. The expected 2028 sundown of the Space-Based Space Surveillance (SBSS) Block 10 satellite puts a clock on the gap.⁶⁵

Reversible counter-electronic-warfare at scale and speed. CCS Block 10.2 and Meadowlands together make up the offensive radio-frequency toolkit; the doctrinal question is whether they are rehearsed at operational speed for symmetric coercion in a Pacific scenario.⁷¹⁷³ The cool war is won at the resilience-and-attribution layer, not at the strike layer. Counter-EW is the credible reversible cost the US can threaten in response to Chinese harassment.

Commercial-operator cyber baseline. The Starshield “unique terms and conditions” framework is the contractual answer, but the cyber posture, cleared-personnel access, and crisis-time legal status of every commercial system Beijing might target are gating items, not nice-to-haves. The first casualty of underinvestment here is the speed of decision-relevant attribution.

Navigation alternatives below the GNSS layer. Multi-system receivers that can fall back from one constellation to another, military-only M-code GPS signals, terrestrial fallbacks like enhanced long-range navigation (eLORAN), optical timing, and inertial backup — the integration of China's BeiDou into US navigation systems is itself a vulnerability, as A Signal Point of Failure argues, and the answer is layered redundancy.⁴⁵

Tactically Responsive Space as a behavior modifier. Victus Nox proved 27-hour tasking; Victus Haze takes the next step, demonstrating on-orbit maneuver with simultaneous surveillance collection. The doctrinal point is simple: the option to put eyes on a target within hours changes Beijing's calculation about what behavior it can hide.⁵⁹⁶⁰

Coalition space-situational-awareness sharing. The AUKUS architecture behind DARC is the model. The cheapest deterrent is denying Beijing a clean cover story; coalition SSA raises the political price of any aggressive action by making attribution shareable at unclassified level across capitals.⁶⁴

Investments that don't

Kinetic anti-satellite capability the US does not intend to use. The 2022 unilateral moratorium is the right strategic posture; building a capability that would violate it would be self-defeating. Kinetic ASAT remains the floor on what the cool war cannot tolerate becoming, not a wartime tool.

Large new exquisite-few GEO systems. Building a small number of expensive, hard-to-replace satellites is structurally vulnerable in a sustained-pressure environment; the proliferated layer is the architectural answer. Investments that re-create single points of failure run against the lesson of the last decade.

What changes if Beijing defects from the ITU

The capability-side change is small. The ITU does not physically gate any Chinese system, and the Resolution 35 milestone mechanic has not yet bitten anyone. The political change is large, because defection removes the rhetorical floor on harassment. The planning implication is that legitimacy infrastructure — coalition norms work, allied operator agreements, SSA sharing — is itself a capability investment. The cheapest deterrent is denying Beijing a clean cover story; the Young and Thadani CSIS analysis argues this is exactly the terrain where Western disorganization is most costly.¹¹²

Methodology, limits, and known gaps

A capability-centric synthesis is a snapshot of a moving picture. This section makes the methodology explicit and lists the places where the open-source record is genuinely thin.

What this is

An open-source analytic synthesis of the US–China counterspace dynamic in low Earth orbit, using the deniability-equilibrium frame as the organizing lens. The argument is built from three layers of source material: the US professional military education ecosystem (West Point's Modern War Institute, Air University's journals and the China Aerospace Studies Institute, the National Defense University's Joint Force Quarterly, and Army War College publications); the analytic infrastructure (CSIS Aerospace Security Project, the Secure World Foundation, RAND, the Aerospace Corporation, IISS); and primary documents (USSF and USSPACECOM doctrine, US–China Economic and Security Review Commission reports, ITU instruments, NASA Orbital Debris Program Office bulletins, ESA's annual Space Environment Report) and trade reporting (SpaceNews, Breaking Defense, Aviation Week, Air & Space Forces Magazine, The Space Review). All sources cited were retrieved or surfaced through targeted research; none are fabricated. Quantitative claims are sourced where the open record supports them; qualitative judgments are framed as analytic synthesis rather than citation.

What this is not

Not a prediction. Not a probabilistic forecast of any specific incident. Not a description of any classified system, plan, or capability. Not an endorsement of any side's framing — the “PRC voice” and “US voice” here are stylized for analytic clarity, not argued for.

Known gaps in the open-source record

The Defense Intelligence Agency's Challenges to Security in Space document is three years stale. The most recent confirmed public edition is April 2022; no 2024 or 2025 successor has surfaced. The headline US intelligence-community public threat-survey predates the 2024 and 2025 SWF debris counts, the April 2024 PLA reorganization, and the Chinese geostationary jamming-experiment satellite.²⁰

No publicly documented case of the ITU enforcing Resolution 35 since 2019. The earliest hard 10% milestone deadlines for major post-2020 mega-constellation filings only fall due around 2027–2029. The teeth of the regime have not yet been tested. Whether the ITU's Radiocommunication Bureau will actually apply MIFR reductions against a defecting state actor is the principal open analytic question for the legitimacy layer.

Quantified attribution timelines. Open-source qualitative judgments from the Aerospace Corporation, CSIS, and SWF describe attribution as “hours to weeks” for jamming and “weeks to never” for cyber. No single open-source benchmark study gives numeric confidence intervals; the field consensus is qualitative.

Specific operational depth of Chinese offensive cyber against US and allied space ground segments. Doctrine-level claims are clear (Burke 2023; CASI's commander-level briefs). Demonstrated operations are largely inferred from PLA academy texts and exercise reporting rather than from disclosed incidents.⁸

The Russian Cosmos 2553 nuclear-ASAT story. Referenced obliquely in the CSIS Space Threat Assessment 2025; no public Aerospace, CSIS, or RAND piece dissects it openly. This is a known unknown that the US intelligence community has acknowledged in unclassified testimony but has not detailed in open report form.¹⁰

The current text of Joint Publication 3-14 (Space Operations). The Joint Staff doctrine library lists Change 1, October 2020 as the operative version, but a current full PDF was not directly retrieved during this research pass. Specific cross-domain response language was therefore not quoted verbatim. The 2025 USSF Space Warfighting framework and Service Doctrine Publications 3-0 and 3-100 supply attribution-supporting language but do not themselves set strategic-response thresholds.⁹⁴⁹⁵

What would change this analysis

Four developments, none imminent on the open-source record as of May 2026, would force a meaningful revision. A documented enforcement action by the ITU's Radiocommunication Bureau against Guowang or Qianfan under Resolution 35 would test the legitimacy-layer hypothesis directly. A confirmed open-source attribution of a discrete close-approach or jamming incident to a named PLA unit would reset the attribution-timeline picture. A public Chinese commitment to the moratorium against destructive direct-ascent ASAT testing would invalidate the framing that the principal counterspace actors are outside the norm. And a documented commercial-operator wartime decision in a Pacific scenario would put the Starshield contractual-control framework to its first real test.

A note on the “cool war” label

The phrase “cool war” is not a term of art in the West Point / Air University / NDU literature. The closest neighbors are “gray zone,” “sub-threshold,” “competition continuum,” “orbital security dilemma,” and “competitive endurance.” The choice to coin a unified label reflects an analytic judgment that the underlying capability dynamic is more coherent than any of those individually-bounded vocabularies. Where this analysis quotes the canonical PME literature, it uses the original terms.