Three billion people still lack reliable internet access. Millions more in rural areas of developed countries struggle with slow, expensive connections. For them, satellite internet isn’t a luxury—it’s potentially their only option.
But here’s what most people don’t realize: not all satellite internet is created equal. The altitude at which satellites orbit fundamentally determines whether your video call feels smooth or awkward, whether gaming is possible or frustrating, and whether the service will cost hundreds or thousands of dollars per month.
This isn’t a matter of engineering preference or company strategy. It’s physics—specifically, the unavoidable trade-offs between orbital mechanics, signal propagation speed, and Earth’s curvature.
Let’s explore why altitude matters so much, and what it means for the future of global connectivity.
The Three Altitude Tiers: A Quick Overview
Satellite internet operates at three distinct altitude bands, each with fundamentally different characteristics:
Low Earth Orbit (LEO): 340–550 km altitude. Fast, requires thousands of satellites. Think Starlink.
Medium Earth Orbit (MEO): Around 8,000 km altitude. Moderate speed, needs hundreds of satellites. Think Blue Origin’s TeraWave.
Geostationary Orbit (GEO): 35,786 km altitude. High latency, only needs three satellites for global coverage. Traditional satellite internet.
The differences between these aren’t just technical details—they create entirely different user experiences.
The Fundamental Physics Trade-Off
Imagine you’re trying to build a network of lighthouses to guide ships along a coast. You have two options:
Option A: Build many small lighthouses close to sea level. Each lighthouse only illuminates a short stretch of coastline, so you need hundreds of them. But ships can see the light quickly because it’s close.
Option B: Build fewer, taller lighthouses on high cliffs. Each one illuminates a much longer stretch of coast, so you need far fewer. But the light takes longer to reach ships because it has to travel farther.
This is exactly the trade-off satellite internet providers face. The physics is inescapable.
Why Altitude Determines Latency
Latency—the delay between sending a request and getting a response—matters enormously for internet quality. Here’s why altitude directly controls latency:
Radio waves (which carry your internet data) travel at the speed of light: approximately 300,000 kilometers per second. That sounds instantaneous, but distance still matters.
For a LEO satellite at 550 km altitude:
- Signal travels 550 km up to the satellite
- Satellite relays it (possibly through other satellites)
- Signal travels 550 km back down
- Round trip minimum: ~1,100 km
- At light speed: roughly 3.7 milliseconds, plus processing time
- Real-world latency: 20–40 milliseconds
For a MEO satellite at 8,000 km altitude:
- Signal travels 8,000 km up
- Signal travels 8,000 km back down
- Round trip minimum: ~16,000 km
- At light speed: roughly 53 milliseconds, plus processing time
- Real-world latency: 80–120 milliseconds
For a GEO satellite at 35,786 km altitude:
- Signal travels 35,786 km up
- Signal travels 35,786 km back down
- Round trip minimum: ~71,572 km
- At light speed: roughly 238 milliseconds, plus processing time
- Real-world latency: 500–700 milliseconds
You can’t engineer your way around this. The speed of light is the speed of light. Higher altitude means longer distance means more latency.
Why Altitude Determines Coverage Area
Earth is a sphere (well, technically an oblate spheroid, but close enough). The higher you are above its surface, the more of it you can see.
A LEO satellite at 550 km can see roughly a 1,000 km diameter circle on Earth’s surface. That sounds like a lot until you realize Earth’s circumference is about 40,000 km. One satellite covers only about 2.5% of Earth’s surface at any moment.
A MEO satellite at 8,000 km can see roughly a 3,000 km diameter circle—about nine times the area of a LEO satellite.
A GEO satellite at 35,786 km can see nearly a third of Earth’s surface. Three strategically placed GEO satellites can cover almost the entire planet (except the poles).
But there’s a catch: satellites don’t stay in one place (except GEO satellites, which orbit at exactly the same speed Earth rotates). LEO satellites scream across the sky, passing over any given spot in just a few minutes before disappearing over the horizon.
The Constellation Size Consequence
To provide continuous coverage to a specific location on Earth, you need enough satellites that as one moves out of range, another is already arriving to take over.
For LEO constellations like Starlink:
- Each satellite covers a small area
- Each satellite moves quickly (orbits Earth in ~90 minutes)
- You need thousands of satellites for global coverage
- Starlink plans to deploy 12,000–42,000 satellites
For MEO constellations like TeraWave:
- Each satellite covers a larger area
- Satellites move more slowly (orbit takes several hours)
- You need hundreds, not thousands
- TeraWave aims for a constellation of ~860 satellites
For GEO satellites:
- Each satellite covers a huge area
- Satellites appear stationary from Earth (they orbit at exactly Earth’s rotation speed)
- You only need 3–4 satellites for global coverage
The economics are obvious: launching thousands of satellites costs vastly more than launching hundreds. But there’s a paradox here—more on that in a moment.
The LEO Approach: Starlink’s Strategy
SpaceX’s Starlink has become synonymous with satellite internet, and for good reason. They’ve deployed over 5,000 satellites and serve millions of customers worldwide.
The Advantages
Low latency: At 20–40 milliseconds, Starlink’s latency is comparable to good cable internet and better than many rural DSL connections. You can video call without awkward delays. You can game competitively (though not at esports levels). The internet feels normal.
Redundancy: With thousands of satellites, if one fails, dozens of others can pick up the slack. There’s no single point of failure.
Speed: Current Starlink speeds range from 50–200 Mbps, with some users seeing bursts above 300 Mbps. That’s genuinely fast by any standard.
The Challenges
Massive deployment costs: Launching thousands of satellites is expensive, even with SpaceX’s cost-effective reusable rockets.
Orbital congestion: LEO is getting crowded. All those satellites need precise coordination to avoid collisions and to manage “handoffs” as satellites move in and out of range.
Limited lifespan: LEO satellites experience atmospheric drag. They need periodic boosts to maintain altitude or they’ll eventually burn up on reentry. Starlink satellites last about 5 years, requiring constant replacement.
Terminal cost: The user equipment (the dish you mount on your roof) costs several hundred dollars because it needs sophisticated phased-array antenna technology to track fast-moving satellites.
The MEO Approach: Blue Origin’s TeraWave
Blue Origin recently announced TeraWave, a MEO constellation designed to compete with Starlink using a fundamentally different approach.
The Advantages
Fewer satellites needed: Around 860 satellites versus Starlink’s tens of thousands. Lower deployment cost and less orbital congestion.
Larger coverage per satellite: Each satellite covers more area, requiring less frequent handoffs between satellites.
Longer lifespan: Less atmospheric drag at higher altitudes means satellites can operate for 10+ years without needing replacement as frequently.
Higher bandwidth per satellite: TeraWave promises 6 Tbps (terabits per second) per satellite using advanced Ka-band frequencies and optical inter-satellite links.
The Challenges
Higher latency: At 80–120 milliseconds, TeraWave’s latency is acceptable for most uses but noticeable for real-time applications. Video calls work fine. Gaming is possible but not ideal. Virtual reality or remote surgery would be problematic.
Less redundancy: Fewer satellites means fewer backups if one fails. Service reliability depends on each satellite working correctly.
More complex satellites: Higher bandwidth and wider coverage require more sophisticated, expensive satellites. Each one costs more, even though you need fewer.
The Coordination Challenge: A Network in Motion
Here’s something most people don’t think about: satellite internet requires solving distributed systems problems while everything is moving at thousands of kilometers per hour.
Handoff Management
When you’re streaming a video via satellite internet, you’re not talking to a single satellite for the entire session. As satellites move, your connection must seamlessly transfer from one satellite to another—sometimes every few minutes with LEO systems.
This is similar to how your cellphone hands off between cell towers as you drive, but more complex because satellites move much faster and cover much larger areas.
The system must:
- Predict when a satellite will move out of range
- Identify which satellite will be in position next
- Transfer your connection without dropping packets
- Do all this while you’re in the middle of a Zoom call or online game
Beam Forming and Coverage Cells
Modern satellite internet doesn’t just broadcast in all directions. Satellites use phased-array antennas that create focused beams—like spotlights—that can be electronically steered without moving the antenna physically.
Each satellite creates multiple coverage “cells,” similar to how cell towers create cellular coverage areas. This allows the satellite to serve many users simultaneously by directing different beams at different locations.
Starlink satellites can create hundreds of individual beams. TeraWave satellites, being higher up with more advanced technology, aim for even more sophisticated beam-forming.
Inter-Satellite Links: The Space Mesh Network
Earlier satellite internet systems sent data up to a satellite, which immediately relayed it back down to a ground station. This created a problem: if you’re in the middle of the ocean, the nearest ground station might be thousands of kilometers away, and data would need to bounce up and down multiple times.
Modern systems solve this with optical inter-satellite links—essentially, fiber optic cables in space. Satellites talk to each other using laser beams, creating a mesh network that routes data through space before sending it back to Earth at the optimal location.
This is technically brilliant but fiendishly complex. Imagine maintaining a network where all the routers are moving at 27,000 km/h (for LEO) and the connections are laser beams that must stay aligned within microradians.
The Economics: Why LEO Might Win Despite Higher Costs
Here’s the paradox: LEO systems cost more to deploy (thousands of satellites versus hundreds), but they might win the market anyway.
Why?
Lower latency enables more use cases: The 20–40 millisecond latency of LEO supports applications that MEO’s 80–120 millisecond latency struggles with. Cloud gaming, VR, real-time collaboration tools, and other latency-sensitive applications work much better on LEO.
Manufacturing at scale reduces costs: Building thousands of identical satellites enables mass production techniques. Starlink satellites are relatively simple and produced on assembly lines. Each unit is cheaper than building hundreds of complex MEO satellites.
Reusable launch vehicles: SpaceX can launch its own satellites on its own rockets at cost. Blue Origin can do the same with New Glenn. But SpaceX has been doing it longer and has demonstrated rapid reusability.
First-mover advantage: Starlink already has millions of subscribers and years of operational experience. They’re iterating on constellation design, satellite technology, and ground terminal improvements faster than competitors can get started.
Real-World Impact: Who Benefits?
Rural and Remote Users
For people living outside urban areas with fiber optic infrastructure, satellite internet is transformative. Suddenly, remote work becomes viable. Students can access online education. Telemedicine becomes possible.
The difference between 40 milliseconds (LEO) and 120 milliseconds (MEO) might seem small, but it affects whether video calls feel natural or stilted, whether cloud-based software is responsive or laggy.
Maritime and Aviation
Ships and aircraft need internet for both operational purposes and passenger services. Low latency enables real-time navigation updates, weather data, and communication with shore-based operations.
Airlines are increasingly offering in-flight WiFi. The quality depends heavily on latency—passengers notice when their video call freezes or their webpage takes forever to load.
Disaster Response and Military
When hurricanes, earthquakes, or floods destroy ground infrastructure, satellite internet provides the only connectivity. Response teams need reliable, low-latency connections for coordination.
Military applications have even stricter requirements. Drone operations, battlefield coordination, and intelligence gathering all benefit from lower latency.
Developing Nations
Many countries lack the infrastructure for widespread fiber optic deployment. Building cell towers in remote regions is expensive. Satellite internet could leapfrog traditional infrastructure entirely.
The question is cost. If satellite internet remains expensive, it won’t help the billions who need it most. This is where MEO systems might have an advantage—fewer satellites could mean lower service costs, making internet accessible to more people even if latency is slightly higher.
The Technical Deep Dive: How Fast Can Data Actually Move?
Let’s dig into the actual bandwidth capabilities, because “6 Tbps per satellite” is a marketing number that needs context.
Understanding Throughput vs. User Experience
When TeraWave claims 6 Tbps per satellite, that’s the theoretical maximum throughput across all beams and all users combined. An individual user won’t get anywhere close to that.
Here’s why:
Shared bandwidth: Those 6 Tbps are split among hundreds or thousands of simultaneous users. If 1,000 users are connected to one satellite, average bandwidth per user is 6 Gbps—still phenomenal, but not 6 Tbps.
Coverage area: A MEO satellite covers a huge area. Users at the edge of coverage get weaker signals than users directly beneath the satellite, affecting their individual speeds.
Beam allocation: The satellite creates many beams, each serving a different geographic area. Bandwidth is divided among these beams based on demand.
Overhead: Protocol overhead, error correction, and retransmission reduce effective throughput.
Real-world user speeds will likely be in the 100 Mbps to 1 Gbps range, depending on congestion and location—excellent by current satellite internet standards, but far from 6 Tbps.
Frequency Bands and Signal Propagation
Satellite internet uses different frequency bands, each with trade-offs:
Ku-band (12–18 GHz): Traditional satellite internet. Decent bandwidth but susceptible to rain fade (heavy rain weakens signals).
Ka-band (26.5–40 GHz): Higher bandwidth, but more affected by weather. Both Starlink and TeraWave use Ka-band for maximum throughput.
V-band (40–75 GHz): Even higher bandwidth potential, but severely affected by atmosphere. Future systems may use this for very high-speed applications.
Higher frequencies carry more data but struggle with atmospheric interference. This is another physics constraint: you can’t have both maximum bandwidth and maximum reliability across all weather conditions.
The Ground Terminal: Your Connection Point
The dish on your roof (or boat, or RV) isn’t a passive antenna. It’s a sophisticated phased-array system that electronically steers its beam to track satellites as they move across the sky.
Starlink’s user terminal contains over 1,000 antenna elements, each independently controlled. By adjusting the phase (timing) of signals across these elements, the dish can point its beam in any direction without physically moving—like how noise-canceling headphones create silence by carefully timing sound waves.
This technology is expensive to manufacture, which is why Starlink terminals cost several hundred dollars. But it’s necessary for tracking fast-moving LEO satellites.
MEO terminals could potentially be simpler because satellites move more slowly and stay in view longer. But they still need phased arrays to maintain connections across the satellite’s wider coverage area.
Looking Forward: What’s Next for Satellite Internet?
The Constellation Race
We’re in the early innings of a satellite internet race:
- Starlink (SpaceX): Operational, 5,000+ satellites, millions of users
- TeraWave (Blue Origin): Announced, aiming for launch in coming years
- Project Kuiper (Amazon): In development, 3,236 planned satellites
- OneWeb: Operational, 600+ satellites, focusing on enterprise and government
Each system makes different trade-offs. Starlink prioritizes latency. TeraWave aims for fewer satellites with higher individual capacity. OneWeb focuses on B2B markets rather than consumers.
The Orbital Debris Problem
Here’s an uncomfortable truth: every satellite eventually becomes debris. LEO satellites have shorter lifespans and burn up on reentry, which is environmentally preferable to leaving junk in orbit. But the sheer number of LEO satellites raises collision risks.
Space agencies worry about Kessler Syndrome—a cascade where one collision creates debris that causes more collisions, eventually rendering certain orbits unusable. With tens of thousands of satellites planned, orbital management becomes critical.
MEO systems have fewer satellites but they stay up longer. Each failure leaves debris in an orbital region that’s harder to clean up than LEO.
This isn’t just an environmental concern—it’s an existential risk to the satellite internet industry itself.
The Regulatory Landscape
Radio spectrum is a limited resource. National and international bodies regulate who can use which frequencies to avoid interference. As more companies launch satellites, spectrum allocation becomes contentious.
Additionally, countries have different regulations about operating satellite internet within their borders. Some nations require local data storage or government access to communications. These regulatory challenges may limit truly global coverage.
The Technology Evolution
Both LEO and MEO systems will improve:
Optical inter-satellite links will become standard, reducing reliance on ground stations and improving routing efficiency.
Laser-based user terminals could eventually replace phased arrays, dramatically reducing terminal cost and size.
AI-based beam management could optimize coverage in real-time based on user demand and satellite position.
Higher frequency bands (V-band and beyond) could enable multi-gigabit speeds for individual users, though weather sensitivity remains a challenge.
The fundamental physics won’t change—altitude will always determine latency and coverage. But engineering improvements can optimize within those constraints.
The Bigger Picture: A Second Internet Infrastructure Layer
What we’re witnessing is the creation of a parallel internet infrastructure that doesn’t depend on terrestrial cables, fiber optics, or cell towers.
This has profound implications:
Resilience: Natural disasters, wars, and accidents can sever ground-based infrastructure. Satellite internet provides backup connectivity.
Global equity: Rural and remote populations gain access comparable to urban fiber—not quite as good, but close enough to be life-changing.
Competitive pressure: Terrestrial ISPs face competition that might drive down prices and improve service, especially in underserved areas.
Geopolitical implications: Countries without advanced terrestrial infrastructure can access global internet without massive ground investments. But this also means potential challenges to information control that some governments maintain.
Making Sense of the Altitude Choice
So which is better—LEO or MEO?
The honest answer: it depends on priorities.
Choose LEO (Starlink) if:
- Latency matters for your use case (gaming, video calls, cloud software)
- You value service availability and redundancy
- You’re willing to pay current market prices
- You’re in an area with existing coverage
Choose MEO (TeraWave, when available) if:
- Basic internet access is your priority over ultra-low latency
- You value potentially lower long-term costs
- Your use case tolerates 80–120 ms latency
- You’re in regions where MEO coverage arrives first
Stick with GEO (traditional satellite) if:
- You’re in extremely remote locations not covered by newer systems
- Latency doesn’t matter (email, basic web browsing)
- It’s your only option
For most people in most situations, LEO systems like Starlink currently offer the best balance of speed, latency, and availability. But MEO systems like TeraWave represent a legitimate alternative approach that could win in markets where lower costs matter more than ultra-low latency.
The Key Takeaway
Satellite internet altitude isn’t a trivial technical detail—it’s a fundamental constraint that shapes everything about the service. Physics dictates that you must choose between coverage efficiency and latency. You cannot have both.
LEO systems choose low latency at the cost of needing thousands of satellites. MEO systems choose fewer satellites at the cost of higher latency. Neither approach is wrong; they optimize for different priorities.
The exciting part is that we’re getting both. Competition between these approaches will drive innovation, reduce costs, and ultimately bring internet connectivity to billions of people who currently lack it.
Understanding the physics behind these trade-offs helps you make informed decisions about which service to choose—and appreciate the extraordinary engineering required to build an internet infrastructure in space.
The next time you connect to satellite internet from a remote cabin, a ship in the middle of the ocean, or anywhere else far from fiber optic cables, think about the satellites racing overhead at thousands of kilometers per hour, coordinating handoffs, routing data through space, and overcoming the fundamental constraints of physics to bring you that video call or web page.
It’s not magic. It’s physics, engineering, and economics working together to solve one of the biggest challenges in global connectivity.