What Is Low Earth Orbit and Why It Is Transforming Global Connectivity

Low Earth Orbit — the band of space between roughly 160 and 2,000 kilometers above the planet — has become the most commercially contested real estate in the space economy. As of early 2026, tracking databases attribute well over three-quarters of all active satellites to LEO, and the constellations being assembled there have redrawn the economics of global connectivity in under a decade. So what is low earth orbit, and why did an altitude band once reserved for imaging satellites and the International Space Station become the default architecture for satellite broadband?

On this page

• From Niche Orbit to the Industry’s Center of Gravity
• What Is Low Earth Orbit? The Definition and the Physics
• Coverage Economics: Why One Satellite Is Never Enough
• Why Low Earth Orbit Changed the Latency Equation
• The Launch Cost Collapse That Made Constellations Viable
• Phased Arrays and Optical Mesh: The Hardware That Made LEO Usable
• LEO vs MEO vs GEO: Where Each Orbit Still Wins
• Industry Implications
• What to Watch

The answer is not a single breakthrough but a compounding set of them: reusable launch collapsed the cost of mass deployment, phased array antennas made tracking fast-moving satellites practical, and inter-satellite links turned swarms of spacecraft into coherent networks. This explainer covers the definition and physics professionals actually need, the constellation economics that follow directly from them, and why LEO — not GEO, and not terrestrial expansion alone — is where the next decade of connectivity infrastructure is being built.

From Niche Orbit to the Industry’s Center of Gravity

Two numbers frame the shift. In 2018, satellite tracking registries counted roughly 2,000 active spacecraft across all orbits. By late 2025 that figure had grown past 11,000, with the overwhelming majority operating in LEO — driven principally by SpaceX, whose Starlink network alone accounts for more than 7,000 operational satellites per public tracking data as of early 2026.

The commercial validation arrived just as fast. Starlink publicly confirmed passing 4 million subscribers in late 2024, with analyst estimates for 2026 running meaningfully higher. OneWeb completed its first-generation constellation and folded into Eutelsat Group. Amazon moved Project Kuiper from license to production launches. [INTERNAL LINK: SpaceX Starlink constellation analysis → full Starlink company profile on leoinsider.com] Capital, spectrum filings and regulatory attention have all followed the satellites down to LEO — which is why understanding the orbit itself is now table stakes for anyone evaluating the sector.

What Is Low Earth Orbit? The Definition and the Physics

By convention, low earth orbit covers altitudes from about 160 km to 2,000 km. Below 160 km, atmospheric drag pulls a satellite down within days; above 2,000 km begins Medium Earth Orbit. The definition matters commercially because everything that makes LEO attractive — and everything that makes it operationally hard — follows from that altitude range.

Orbital mechanics is unforgiving about the terms. A satellite at LEO altitude must travel at roughly 7.8 km/s to stay in orbit, completing a full circuit of the planet every 90 to 110 minutes — 15 to 16 orbits per day. From any point on the ground, an individual LEO satellite is visible for only a few minutes per pass. The commercially popular shells cluster at specific altitudes: Starlink operates primarily around 550 km, OneWeb’s first generation at 1,200 km, and most Earth-observation systems in sun-synchronous orbits between 500 and 800 km.

Drag, Decay and the Self-Cleaning Orbit

Below roughly 500 km, residual atmosphere continuously slows spacecraft, and an unpowered satellite deorbits within a few years — a property operators now market as a debris-mitigation feature. Regulation has hardened around the same idea: the FCC’s post-mission disposal rule requires LEO satellites licensed under it to deorbit within five years of end of mission, a sharp tightening from the old 25-year guideline. [EXTERNAL LINK: FCC orbital debris mitigation order → FCC.gov rulemaking documents] Higher LEO shells above 1,000 km do not self-clean on useful timescales, which is why debris policy treats them differently.

Coverage Economics: Why One Satellite Is Never Enough

A single satellite in geostationary orbit sees roughly a third of the planet and appears fixed in the sky; three or four deliver near-global coverage. A single LEO satellite, by contrast, illuminates a footprint a few hundred kilometers across — and that footprint is moving at nearly 28,000 km/h. Continuous service from LEO therefore requires a constellation: enough satellites, in enough orbital planes, that one is always overhead.

Iridium proved the minimum viable answer decades ago with 66 cross-linked satellites for narrowband voice. Broadband raises the bar by an order of magnitude: meaningful capacity per user pushes constellation sizes into the hundreds or thousands. That single fact restructures the industry around serial manufacturing — SpaceX has publicly discussed build rates measured in satellites per day, not per year — and around a permanent replacement cycle, since LEO spacecraft are designed for roughly five-to-seven-year service lives. Constellation operators are not running one capital project; they are running a production line forever. That is the scale barrier that keeps the field to a handful of credible players.

Why Low Earth Orbit Changed the Latency Equation

Latency is where LEO stopped being a compromise and became a competitor. Geostationary distance imposes physics that no engineering can remove: a round trip to 35,786 km costs roughly 480 ms before any network processing. Consumer GEO internet in practice delivers 600 ms or more — workable for browsing, hostile to video calls, VPNs and real-time applications.

From a 550 km shell, the same physics costs about 7 ms, and operator-reported real-world figures for LEO broadband cluster between 25 and 60 ms — inside the range of terrestrial fixed access in much of the world. [INTERNAL LINK: LEO vs GEO latency comparison → technical latency analysis on leoinsider.com] The consequence for the market is structural: LEO connectivity competes for enterprise WAN traffic, maritime and aviation backbones, and even latency-sensitive financial links, rather than serving only as coverage of last resort. Optical inter-satellite links reinforce the advantage by routing traffic through the mesh in orbit, reducing dependence on local ground infrastructure.

The Launch Cost Collapse That Made Constellations Viable

None of this is conceptually new. Iridium, Globalstar and Teledesic all designed LEO constellations in the 1990s — and the first two emerged from bankruptcy while the third never launched. The physics worked; the launch economics did not. Deploying hundreds of satellites at expendable-rocket prices broke every business plan that tried it.

Reusability changed the input costs. Falcon 9 boosters now routinely fly more than 20 missions, and industry estimates put per-kilogram costs to LEO an order of magnitude below early-2000s expendable pricing — with SpaceX’s internal cost for Starlink launches lower still, a structural advantage of owning the rocket. Dedicated rideshare programs did the same for smaller operators. The next inflection is pending: Starship is designed explicitly for mass constellation deployment, and each step-change in lift capacity re-prices what a constellation costs to build and replenish.

Phased Arrays and Optical Mesh: The Hardware That Made LEO Usable

Cheap launch solved deployment; it did not solve the user side. A LEO satellite crosses the visible sky in minutes, and historically that meant motorized tracking dishes — acceptable for a teleport, absurd for a household or a vessel. Electronically steered phased array antennas removed that constraint: flat panels that switch beams in milliseconds with no moving parts, handing off between satellites invisibly. The cost curve has been steep in the right direction — early flat-panel terminals were multi-thousand-dollar hardware, while standard consumer kits now retail in the low hundreds of dollars, with operators publicly acknowledging they subsidize hardware to win subscribers. Terminal economics, not satellite economics, remain the gating factor for mass-market adoption.

The space segment matured in parallel. Optical inter-satellite links — laser terminals operating at up to 25 Gbps per link — let traffic ride the constellation as a mesh network, touching ground only near its destination. The operational consequences are significant: coverage over oceans and polar routes without local gateways, service into jurisdictions where ground infrastructure is impractical, and long-haul paths that network modeling suggests can compete with terrestrial fiber on some routes, since light travels faster in vacuum than in glass.

The pattern worth internalizing: reusable launch, phased arrays and optical mesh all reached commercial maturity within the same decade. Remove any one of the three and the LEO broadband model reverts to its 1990s outcome.

LEO vs MEO vs GEO: Where Each Orbit Still Wins

LEO’s rise does not make the other orbits obsolete — it re-sorts their roles. The comparison professionals actually use looks like this. [INTERNAL LINK: LEO vs GEO vs MEO satellites → the complete industry orbital comparison]

Orbit	Altitude	Typical round-trip latency	Satellites for global coverage	Best commercial fit
LEO	160–2,000 km	25–60 ms	Hundreds to thousands	Broadband, direct-to-device, Earth observation
MEO	2,000–35,786 km	~130–150 ms	Tens	Backhaul, GNSS/navigation
GEO	35,786 km	~600 ms	3–4	Broadcast, fixed wide-area capacity

GEO retains genuine advantages — persistent coverage of a fixed region and unmatched simplicity for broadcast — and MEO remains the home of navigation constellations and niche low-latency backhaul. The strategic pattern of 2026 is multi-orbit: incumbent operators increasingly sell LEO capacity alongside their own GEO or MEO assets rather than against them.

Frequently Asked Questions

What altitude range counts as low earth orbit?

By ITU and industry convention, LEO spans roughly 160 km to 2,000 km above Earth. Below 160 km atmospheric drag deorbits spacecraft within days; above 2,000 km begins Medium Earth Orbit. Most commercial broadband constellations operate between 500 and 1,200 km.

Why do LEO systems need hundreds of satellites when GEO needs three?

A GEO satellite at 35,786 km sees a third of the planet and appears stationary. A LEO satellite’s footprint is a few hundred kilometers wide and crosses the sky in minutes, so continuous coverage requires enough satellites that one is always overhead — 66 for Iridium’s narrowband service, thousands for meaningful broadband capacity.

How long do LEO satellites stay in orbit?

Design lives run about five to seven years for broadband constellations. Below ~500 km, drag deorbits dead satellites naturally within a few years; FCC rules now require post-mission disposal within five years for satellites licensed under them. Higher shells above 1,000 km do not self-clean on useful timescales.

Is LEO connectivity proven enough for enterprise use?

Yes, with qualifications. Real-world latency of 25–60 ms and maturing SLA-backed offerings have made LEO a credible enterprise WAN and backup option, and maritime and aviation adoption is well documented. Buyers should still evaluate per-site throughput commitments and terminal logistics rather than headline speeds.

What should investors look at when evaluating a LEO operator?

Three fundamentals: launch access and its cost trajectory, the replacement-cycle capex implied by constellation size and satellite lifespan, and subscriber economics per segment. Headline satellite counts are a vanity metric without the manufacturing and launch economics underneath them.