Part 0: Read This First | Part 1: Undersea Cable Empire | PART 2: THE SATELLITE SOVEREIGNTY RACE | Part 3: DNS Dictatorship | Part 4: Payment Rails | Part 5: The Cloud Is Someone's Computer | Part 6: Credential Wars
Part 2: The Satellite Sovereignty Race
When Cables Can Be Cut, Satellites Become the Backup—But Who Controls Orbit?
February 24, 2022. Russia invades Ukraine. Within hours, Ukrainian military communications face disruption as cell towers are destroyed and fiber networks are cut. The Ukrainian government makes an emergency request: activate Starlink. Elon Musk responds on Twitter. Within 48 hours, Starlink terminals arrive in Ukraine. Within a week, thousands are operational. Ukrainian forces use them to coordinate drone strikes, maintain command networks, and broadcast from the front lines—all while Russian jamming efforts prove largely ineffective against the satellite network. For the first time in modern warfare, a private satellite constellation became critical military infrastructure. Not government-owned. Not treaty-regulated. A commercial system owned by one company, controlled by one person, deployed in an active war zone. The implications were immediate: satellites aren't just backup internet for rural areas. They're infrastructure that bypasses everything terrestrial—cables, cell towers, fiber networks, government control. If your ground infrastructure gets cut, satellites keep you connected. If your adversary controls ground networks, satellites give you independence. The cable empire is vulnerable to cutting. The satellite empire is vulnerable to something else: orbital competition. And that race is already underway. Starlink has 7,000+ satellites in orbit as of January 2026, with plans for 42,000. China is building the Guowang (GW) constellation—13,000+ satellites planned, state-funded, explicitly designed as a Starlink alternative. Russia, the EU, India, and others are launching their own systems. Low Earth Orbit is becoming the most contested infrastructure domain on the planet. This isn't about internet access. It's about who controls the high ground when ground networks fail. Welcome to the satellite sovereignty race.
The LEO Revolution: Why Low Earth Orbit Changes Everything
For decades, satellite internet meant geostationary satellites—massive spacecraft sitting 35,786 kilometers above Earth, appearing stationary from the ground. They worked, but they were slow (500+ millisecond latency), expensive ($500+ million per satellite), and limited in capacity (a few gigabits per second total).
Low Earth Orbit (LEO) satellites—operating at 500-1,200 km altitude—change the entire equation:
- Lower latency: 20-40 milliseconds (comparable to fiber optic cables)
- Cheaper to build: $250,000-1 million per satellite (vs $500 million for geostationary)
- Higher capacity: Entire constellations can deliver terabits per second
- Global coverage: Polar orbits cover the entire planet, including oceans and polar regions
- Harder to jam: Signal hops between satellites, making traditional jamming ineffective
The tradeoff: LEO satellites have shorter lifespans (5-7 years vs 15+ for geostationary), require massive constellations (thousands instead of dozens), and create orbital congestion. But the strategic advantages outweigh these costs.
Why LEO matters for infrastructure competition:
LEO satellites can bypass every terrestrial chokepoint we mapped in Part 1. Undersea cables concentrate through Suez, Malacca, Taiwan Strait—all vulnerable. LEO satellites orbit above all of it. If cables get cut, satellites provide redundancy. If landing points get seized, satellites don't need landing rights in the same way (ground stations can be mobile, ship-based, or rapidly deployed).
This makes LEO satellite constellations the ultimate infrastructure hedge—insurance against terrestrial network failure.
TRADITIONAL GEOSTATIONARY:
• Altitude: 35,786 km
• Latency: 500-700 ms
• Cost per satellite: $500M+
• Lifespan: 15+ years
• Global fleet: ~560 satellites
• Use case: Broadcast TV, legacy internet
MODERN LEO CONSTELLATIONS:
• Altitude: 500-1,200 km
• Latency: 20-40 ms
• Cost per satellite: $250K-1M
• Lifespan: 5-7 years
• Global fleet: 10,000+ satellites (2026)
• Projected fleet: 50,000+ by 2030
• Use case: Broadband, military comms, IoT
THE SHIFT:
In 2019: ~2,000 total satellites in orbit (all types)
In 2026: ~12,000+ satellites (mostly LEO)
By 2030: Projected 50,000+ (LEO explosion)
WHY THIS MATTERS:
LEO makes satellite internet viable at scale.
And whoever owns the constellation owns the bypass
to terrestrial infrastructure control.
Starlink's First-Mover Advantage: 7,000 Satellites and Counting
SpaceX's Starlink is the dominant LEO constellation by a massive margin. As of January 2026:
- Satellites deployed: ~7,000 operational
- Approved for deployment: 12,000 (FCC authorization)
- Additional application pending: 30,000 more (42,000 total planned)
- Launch rate: 40-60 satellites per week (using Falcon 9 and Starship)
- Subscribers: 3+ million globally (as of late 2025)
- Coverage: 60+ countries, expanding to 100+ by end of 2026
Starlink isn't just ahead—it's lapping the competition. The next-largest constellation (OneWeb) has ~650 satellites. Amazon's Project Kuiper has launched test satellites but hasn't begun mass deployment. China's GW constellation is in early stages.
Why Starlink's lead matters:
1. Network Effects in Orbit
More satellites = better coverage, lower latency, higher capacity. A 1,000-satellite constellation provides basic coverage. A 7,000-satellite constellation provides redundancy, load balancing, and resilience. A 42,000-satellite constellation would be nearly impossible to degrade through selective attacks—you'd need to eliminate thousands of satellites to significantly impact service.
2. Launch Cost Advantage
SpaceX owns the rockets (Falcon 9, Starship). Internal launch costs are estimated at $15-30 million per 60-satellite batch—far cheaper than competitors paying $50-100 million for third-party launches. This vertical integration makes Starlink's economics unbeatable.
3. Dual-Use Infrastructure
Starlink is officially a commercial system. But it has explicit contracts with the US Department of Defense, National Reconnaissance Office, and Space Force. The distinction between "civilian" and "military" infrastructure is increasingly meaningless.
Military uses of Starlink (confirmed or reported):
- Ukraine military communications (2022-present)
- US military exercises (Arctic, Pacific, contested regions)
- Classified government networks (Starshield program—dedicated military variant)
- Intelligence agencies (surveillance data relay)
Starlink operates as commercial infrastructure with a military overlay. This gives the US military global communication infrastructure that's technically private sector—harder to target legally, diplomatically, or kinetically.
4. Geopolitical Leverage
Starlink can provide or deny service to entire countries. Examples:
- Ukraine (2022): Service activated within 48 hours of request
- Crimea (2023): Service geofenced to prevent Ukrainian use in Russian-controlled territory (Musk decision, controversial)
- Iran (2022): Terminals shipped to support protests, blocked by export controls
- China: Service not available (Chinese government would never allow it)
A private company now has the power to decide which nations get satellite internet access. That's unprecedented.
TOOL 1: CelesTrak
Website: celestrak.org
Tracks satellite orbits in real-time. Search "Starlink" to see all 7,000+ satellites, their orbital paths, altitudes, and positions.
TOOL 2: N2YO Satellite Tracker
Website: n2yo.com
Live tracking of any satellite. Shows current position, next pass over your location, orbital parameters. Track Starlink satellites passing over your city in real-time.
TOOL 3: Starlink Coverage Map
Website: starlink.com/map
Official coverage map showing where Starlink service is available. Notice the gaps—China, Russia, most of Africa (where GW constellation will compete).
EXPERIMENT:
Use CelesTrak to count how many Starlink satellites pass over a contested region (Taiwan Strait, South China Sea, Ukraine) in 24 hours. You'll understand why satellite coverage is so hard to deny.
China's Response: The Guowang (GW) Constellation
China watched Starlink's role in Ukraine and drew a clear conclusion: satellite constellations are strategic infrastructure that cannot be ceded to competitors.
Enter the Guowang (GW) constellation—China's state-backed answer to Starlink.
GW Constellation specs (as of 2026):
- Planned satellites: 12,992 (initial approval), potentially expanding to 30,000+
- Currently deployed: ~300 test satellites
- Operator: China Satellite Network Group (state-owned enterprise, established 2021)
- Orbital shells: 500 km, 600 km, 1,145 km altitudes
- Timeline: Mass deployment began 2024, targeting 5,000+ satellites by 2027
- Funding: State-backed, estimated $100+ billion over 10 years
- Purpose (official): Broadband internet for rural China, Belt & Road countries
- Purpose (strategic): Communication independence, military backup, orbital denial
Why China Is Building GW
1. Communication Sovereignty
China cannot allow critical infrastructure to depend on American satellites. If conflict occurs, Starlink could be denied to China (it already is) or used against Chinese interests (as in Ukraine). GW provides an alternative that China fully controls.
2. Belt & Road Extension
GW will provide satellite internet to Belt & Road countries—Pakistan, much of Africa, Southeast Asia, Central Asia. Just as China built alternative undersea cable routes (PEACE cable), GW extends Chinese infrastructure into space. Countries using GW become dependent on Chinese satellite infrastructure.
3. Military Redundancy
If undersea cables get cut (Taiwan conflict, South China Sea disruption), GW provides communication backup for Chinese military forces, government networks, and critical infrastructure. It's the space layer of China's A2/AD (Anti-Access/Area Denial) strategy.
4. Orbital Denial Strategy
Here's the darker possibility: by filling Low Earth Orbit with 13,000+ satellites, China makes it harder for competitors to deploy additional constellations. Orbital slots are limited. Frequency spectrum is finite. The first movers claim the best positions. If China and the US together deploy 50,000+ satellites, there may not be room for others.
The Launch Challenge
China's biggest constraint: launch capacity. Starlink can launch 60 satellites per Falcon 9 flight at a rate of 50+ flights per year. China's Long March rockets can launch 10-30 satellites per flight at a rate of 20-30 flights per year (for all purposes, not just satellites).
To catch up, China is developing:
- Reusable rockets (Long March 9, testing underway)
- Commercial launch companies (Landspace, iSpace, Galactic Energy—state-supported)
- Rapid launch infrastructure (new launch sites in Hainan, Wenchang)
If China achieves reusable rocket capability comparable to SpaceX by 2027-2028, the GW deployment could accelerate dramatically. If not, the constellation will take 10+ years to complete.
Unlike undersea cables (concentrated through geographic straits), satellites have different chokepoints:
1. ORBITAL SHELLS (Altitude Limits)
• LEO sweet spot: 500-1,200 km (low latency + long enough lifespan)
• Too low (<400 km): Atmospheric drag, satellites decay in months
• Too high (>1,500 km): Longer latency, debris stays in orbit for centuries
• The 500-1,200 km shell can only fit ~50,000-60,000 satellites safely
• Starlink + GW + others will fill this by 2030
2. FREQUENCY SPECTRUM (Radio Limits)
• Ka-band, Ku-band, V-band: Limited spectrum available
• ITU (International Telecom Union) allocates spectrum
• First to file gets priority (hence the deployment race)
• Spectrum conflicts = interference = service degradation
3. LAUNCH CAPACITY (Access Limits)
• Rockets needed: Hundreds of launches to deploy 10,000+ satellites
• Only two entities can do this at scale: SpaceX (US), potentially China by 2028
• Launch capacity = deployment speed = strategic advantage
4. GROUND STATIONS (Control Limits)
• Satellites need ground stations (gateways) to connect to terrestrial internet
• Ground stations need host country approval
• China can deny Starlink ground stations, US can deny GW
• Countries become chokepoints for satellite networks
5. ORBITAL DEBRIS (Kessler Syndrome Risk)
• Each satellite eventually becomes debris
• Collisions create more debris (cascading effect)
• If debris density gets too high, LEO becomes unusable for decades
• This is the ultimate chokepoint: pollute LEO, deny it to everyone
CONCLUSION:
Space chokepoints aren't geographic—they're physical (altitude),
regulatory (spectrum), and existential (debris). And the race to
claim them is happening right now.
The Other Players: OneWeb, Kuiper, and National Systems
Starlink and GW dominate, but they're not alone:
OneWeb (UK/EU/India)
- Satellites deployed: ~650 (constellation complete as of 2023)
- Ownership: UK government, Bharti (India), Eutelsat (France)
- Strategy: Focus on enterprise, government, mobility markets (not residential)
- Status: Operational but small-scale compared to Starlink
Amazon Project Kuiper
- Planned satellites: 3,236
- Currently deployed: 2 test satellites (as of late 2025)
- Launch contracts: $10+ billion committed (ULA, Arianespace, Blue Origin)
- Timeline: Mass deployment starting 2026, service by 2027
- Challenge: Late to market, but Amazon has capital and AWS integration
National/Regional Systems
- Russia (Sfera): 600+ satellites planned, minimal deployment so far
- EU (IRIS²): 170-satellite constellation planned, government-funded, 2027 target
- India (NavIC expansion): Considering LEO constellation for rural connectivity
The pattern: nations recognize satellites as strategic infrastructure and are building sovereign systems. But only the US (Starlink) and China (GW) have the scale to provide global coverage.
Military Control of "Civilian" Systems
Every major satellite constellation has military integration:
Starlink:
- Starshield (classified military variant)
- DoD contracts worth $hundreds of millions
- Used in Ukraine (combat-tested)
- Integration with US Space Force operations
GW Constellation:
- Operated by state enterprise with PLA ties
- Dual-use from inception (civil + military)
- Will support Chinese military operations in Taiwan scenarios, South China Sea
OneWeb:
- UK government ownership stake explicitly for "sovereign communication"
- Contracts with NATO militaries
The fiction of "civilian" satellite infrastructure is dissolving. These are military assets with commercial facades.
STARLINK (SpaceX):
• Estimated total investment: $30+ billion (2020-2026)
• Annual deployment cost: $5+ billion (launches, satellites, ground infrastructure)
• Revenue (2025): $6+ billion (subscriber fees, government contracts)
• Projected revenue (2030): $30+ billion
• Funding: Private investment + customer revenue + government contracts
GW CONSTELLATION (China):
• Estimated total investment: $100+ billion (2021-2030 projection)
• Annual deployment: $10-15 billion (state-funded)
• Revenue model: State subsidy + Belt & Road subscribers + military contracts
• Not profit-driven—strategic infrastructure investment
PROJECT KUIPER (Amazon):
• Estimated investment: $10+ billion committed
• Launch contracts: $10 billion alone
• Timeline: 2026-2029 deployment
• Revenue model: AWS integration, enterprise customers
ONEWEB:
• Total investment: $5+ billion (initial bankruptcy + rescue)
• Current status: Operational but small-scale
GLOBAL SATELLITE INDUSTRY:
• Total LEO constellation investment (2020-2030): $200+ billion projected
• Annual growth: 25-30%
• Launch industry revenue (driven by satellites): $15+ billion/year
THE TAKEAWAY:
Space infrastructure is seeing more investment in 10 years
than the previous 50 years combined. The race is accelerating,
not slowing. By 2030, LEO will be unrecognizable.
Historical Parallel: Air Superiority in WWII
THE SETUP:
At the start of WWII, air power was considered supplementary to ground and naval forces. By 1945, air superiority was recognized as a prerequisite for winning any major engagement.
THE LESSON:
Whoever controlled the air could:
• Disrupt enemy supply lines (strategic bombing)
• Provide reconnaissance (see enemy movements)
• Support ground forces (close air support)
• Deny the same capabilities to enemies
BATTLE OF BRITAIN (1940):
Germany's failure to achieve air superiority over Britain prevented invasion. Control of the air = control of the battlefield below.
PACIFIC THEATER (1942-1945):
US air superiority (carriers, long-range bombers) enabled island-hopping strategy. Japan lost because it couldn't contest airspace.
D-DAY (1944):
Allied invasion succeeded because of air superiority—German forces couldn't move in daylight without being bombed.
THE PRINCIPLE:
"Control the high ground, control the battlefield."
2026 PARALLEL:
Satellites are the new high ground. Whoever controls Low Earth Orbit can:
• Provide communication when ground networks fail
• Conduct reconnaissance (imaging satellites)
• Deny the same to adversaries (anti-satellite weapons, jamming)
• Coordinate forces globally without terrestrial infrastructure
Air superiority determined WWII outcomes.
Orbital superiority may determine 21st-century conflicts.
And the US (Starlink) currently has it. China (GW) is racing to challenge it.
The Alternative Scenario: When Satellites Start Getting Shot Down
TRIGGER EVENT:
A Taiwan crisis escalates. China invades. The US provides military support. Both sides face a problem: communication infrastructure is vulnerable. Undersea cables around Taiwan are cut (see Part 1). Ground networks are jammed or destroyed.
DAY 1: STARLINK BECOMES CRITICAL
• Taiwan military uses Starlink for command networks
• US forces use Starlink for coordination
• Chinese jamming proves ineffective against satellite network
• China faces strategic disadvantage: adversary has space-based communication backup
DAY 3: CHINA RESPONDS
• Anti-satellite (ASAT) missile launched from ground or ship
• One Starlink satellite destroyed over Western Pacific
• Debris field created, but network barely affected (7,000 satellites = redundancy)
• US condemns attack, doesn't escalate (one satellite isn't worth WWIII)
WEEK 1: ESCALATION
• China launches multiple ASATs, destroying 15-20 Starlink satellites
• US destroys Chinese reconnaissance satellites in retaliation
• Both sides target each other's satellite constellations
• GW constellation (partial deployment) suffers heavy losses
• OneWeb, civilian satellites become collateral damage
WEEK 2: KESSLER SYNDROME RISK
• Destroyed satellites create debris fields in LEO
• Debris collides with other satellites (cascading effect)
• Orbital environment degrades rapidly
• Both sides face choice: continue destroying satellites and risk making LEO unusable for decades, or stop
MONTH 1: THE NEW EQUILIBRIUM
• Informal ceasefire on satellite attacks (both sides realize mutual destruction)
• LEO littered with debris, making new launches risky
• Starlink degraded but functional (lost 500 satellites, 6,500 remain)
• GW constellation crippled (lost 60% of deployed satellites)
• Global satellite internet becomes unreliable
• Space industry faces decades of debris cleanup
AFTERMATH:
• LEO becomes contested domain (like airspace in wartime)
• Countries accelerate ground-based laser ASAT systems
• Satellite constellations become legitimate military targets
• The era of "peaceful" space infrastructure ends
• Kessler
THE LESSON:
Satellite constellations provide infrastructure independence—
until they become military targets. Then they're just as vulnerable
as undersea cables. Maybe more so, because debris in space
persists for decades, potentially denying LEO to everyone.
This scenario isn't hypothetical. Both US and China have demonstrated
ASAT capabilities. Russia destroyed its own satellite in 2021 (test),
creating 1,500+ debris pieces. India did the same in 2019.
The weapons exist. The question is when, not if, they're used in conflict.
Conclusion: The High Ground Is Being Claimed
The satellite sovereignty race reveals a fundamental shift in infrastructure strategy: when terrestrial infrastructure can be cut, orbital infrastructure becomes the backup—but orbital infrastructure can also be destroyed, and the debris persists.
The race is accelerating:
- Starlink: 7,000 satellites deployed, 42,000 planned, military-integrated, global coverage
- GW Constellation: 300 deployed, 13,000 planned, state-funded, strategic alternative to US dominance
- Others: OneWeb, Kuiper, national systems—all trying to claim orbital positions before they're gone
By 2030, Low Earth Orbit will host 50,000+ satellites. The first movers (Starlink, GW) will control the best orbital shells, frequency spectrum, and strategic positions. Late movers will face congestion, interference, and debris risk.
This isn't about "internet access for rural areas." It's about:
- Communication independence (bypass terrestrial infrastructure)
- Military redundancy (maintain command networks when ground systems fail)
- Strategic denial (fill orbital slots so competitors can't)
- Geopolitical leverage (provide or deny service to entire regions)
The cable empire is vulnerable to cutting. The satellite empire is vulnerable to shooting. Both are becoming militarized. Both are fragmenting into US-aligned vs. China-aligned systems.
And the high ground—Low Earth Orbit—is being claimed right now, while most people watch TikTok debates and platform regulations.
Space isn't the final frontier. It's the next chokepoint. And the race to control it is already underway.
Next: Part 3 - The DNS Dictatorship (The internet's phonebook is controlled by 13 root servers. 10 are US-controlled. What happens when that system fragments?)
