The International Telecommunication Union in Geneva keeps a ledger of who owns what slice of the sky, and in that ledger sit thousands of satellites that do not exist. They have names, orbital slots, frequency bands, coordination filings, and coverage maps. They have no metal. Radio engineers and space lawyers call them paper satellites, and the earliest filing — sometimes made a decade before any hardware is bent — quietly locks out every operator who tries to come later.
The system works on a rule that sounds innocuous until you follow it to its conclusion: first come, first served. File early with the ITU, defend your paperwork through the coordination process, and the geostationary arc above the equator becomes yours to sit on. Or to sell. Or to sit on and sell.
The 36,000-kilometre parking lot
Geostationary orbit is a single ring, 35,786 kilometres above the equator, where a satellite’s orbital period matches Earth’s rotation and it appears to hover over one spot on the ground. Every direct-to-home TV dish, most weather satellites, and a large share of military and diplomatic communications live there. The ring is roughly 265,000 kilometres in circumference, which sounds enormous until you realise satellites sharing a frequency band typically need to be spaced around two degrees apart to avoid drowning each other out. That leaves approximately 180 usable slots for the whole planet.
The ITU, a United Nations agency founded in 1865 as the International Telegraph Union to sort out telegraph interference between European empires, is the referee. Countries file for a slot and a frequency, the ITU circulates the filing, other administrations raise objections if their own satellites would be affected, and eventually — years later — the filing either matures into a coordinated assignment or lapses.
How a paper satellite is born
The mechanism is simple. An administration — usually a national telecom regulator acting on behalf of a company — submits Advance Publication Information to the ITU describing a satellite it intends to launch. The clock then starts on a bringing-into-use deadline. Under current rules, the operator typically has seven years from the filing date to place an actual working satellite in the slot and broadcast on the notified frequencies for a continuous 90 days.
Seven years is a long time. Long enough to file speculatively, watch the market, and either build something, lease the slot to someone who will, or drag another satellite already in orbit over to the position at the last minute to tick the box. That last trick — flying an existing spacecraft across the arc to briefly occupy a slot and preserve the filing — has a nickname of its own: satellite parking. In 2024, Thaicom completed the relocation of Thaicom-9A to 50.5 degrees East specifically to hold an orbital position, a legal move that is entirely normal in the industry.
The Tongan gambit
The clearest illustration of what paper satellites can do happened in 1988, when the tiny Kingdom of Tonga filed with the ITU for sixteen geostationary slots over the Pacific. Tonga had no satellites, no launch capability, and no immediate need for sixteen orbital positions. What it had was a shrewd American consultant and a reading of ITU rules that suggested a small nation could claim as much of the sky as a large one.
The filing detonated a diplomatic row. Intelsat, the international consortium that then dominated commercial satellite communications, argued Tonga was trying to warehouse spectrum it could never use. After years of coordination the kingdom ended up with several slots, which it leased to operators who did want to broadcast over the Pacific. Tongasat, the company set up to manage the leases, generated substantial revenue for a country whose main exports were squash and vanilla. The episode changed how the ITU wrote its rules, but it did not end the practice — it professionalised it.
The backlog problem
By the early 2010s the ITU’s coordination queue had swollen to tens of thousands of pending filings, most of them for satellites nobody was seriously building. Each filing had to be examined, circulated, and objected to by other administrations, and each one clogged the pipeline for operators who genuinely wanted to fly hardware. The ITU tightened the bringing-into-use rules at World Radiocommunication Conferences in 2012 and 2015, requiring proof that a real satellite had actually operated in the slot rather than merely passed through. Filings that lapsed were supposed to fall out of the register.
They did, sometimes. But new filings kept coming, and the rise of large non-geostationary constellations — Starlink, OneWeb, and their competitors — added a second front. Non-geostationary systems now file for tens of thousands of individual satellites at a time, and the ITU introduced milestone rules requiring operators to deploy specified percentages of a constellation within set timeframes. Miss a milestone and the filing shrinks to whatever has actually been launched.
Why paper still beats metal
A first-mover position in orbit is worth defending because the physics do not permit do-overs. Two satellites cannot share the same slot on the same frequency without interfering, and moving a coordinated slot to accommodate a latecomer requires the incumbent’s consent — consent that usually comes with a price tag. In business and technology, arriving first often shapes the rules that everyone after has to work within. In the geostationary arc that shaping is literal: the ITU’s coordination process treats the earlier filing as the reference point, and the later filer has to prove they will not cause harmful interference to it.
That leverage translates directly into money. Orbital slots over dense TV markets — the Americas at 101 degrees West, Europe at 19.2 degrees East, the Indian Ocean at 68.5 degrees East — command lease values that can run into hundreds of millions of dollars over a satellite’s lifetime. A country that holds the paper on one of those slots holds an asset it can rent to whichever operator can actually build a spacecraft.
The lawyers who read the arc
An entire micro-profession has grown up around ITU filings. Space lawyers at firms in Washington, London, Geneva, and Luxembourg spend their careers tracking coordination triangles, drafting objections, and negotiating the horse-trades that let two administrations share a slot. The international space law framework that governs all of this rests on the 1967 Outer Space Treaty, which forbids any nation from claiming sovereignty over outer space itself, and on the ITU’s Radio Regulations, which do not claim to allocate real estate — only the right to transmit at a given frequency from a given position without interference.
The distinction matters. Nobody owns 19.2 degrees East. What Luxembourg’s SES owns is the ITU-coordinated right to broadcast Ku-band and Ka-band signals from that longitude without another operator drowning them out. Lose the coordination and you lose the business, even if your satellite is still bolted to its solar arrays and drifting on station.
The tension between the treaty language — the moon and other bodies as the province of all mankind — and the practical reality of a queue-based spectrum register is the central puzzle of orbital law. There is no world court for the geostationary arc. There is only the ITU, its filings, its deadlines, and the operators who learn to game them.
What lapses, what stays
The register is not static. Filings expire. Recent rule changes have forced a purge of assignments whose bringing-into-use claims could not be substantiated, and the ITU secretariat now demands documentary evidence — satellite telemetry, transmission logs, coordination confirmations — before recording a slot as brought into use. Filings that were speculative in the 1990s and never matured have been quietly removed. Whether the pace of removals keeps up with the pace of new speculative filings is a running argument at each World Radiocommunication Conference.
Meanwhile the arc keeps filling. Roughly 550 active geostationary satellites orbit Earth today, sharing those 180-odd slots through careful frequency planning. Behind them, in the ITU’s Master International Frequency Register, sit thousands more entries — some tied to hardware being assembled in cleanrooms in Toulouse or El Segundo, some tied to shell companies in jurisdictions that have not built a satellite in decades.
The ledger nobody quite trusts
A geostationary satellite, seen from the ground, is a fixed point of light against the moving stars. Every one of those points began as a filing on a form in Geneva, years before its solar panels unfurled. Some of the filings still open in the register today were made when the operators who submitted them were counting on markets — direct broadcast television in emerging economies, mobile satellite telephony — that turned out very differently than the business plans predicted.
The paperwork outlived the assumptions. It usually does. The next operator who wants a slice of the same sky will open the ITU database, see a filing dated 2003 or 1998 or 1991, and start the slow work of coordinating around a satellite that was never built by a company that may no longer exist — because on the register, in the only ledger that matters, the paper satellite got there first.