How Does High-Speed Worldwide Satellite Internet Work?

Satellite internet is generally one of your last choices – the latency and vulnerability to bad weather make it worse than fiber and cable, but at least it’s not dial-up. However, there are currently several projects, most notably SpaceX’s Starlink, that are aiming not only to make satellite internet faster but to make it available everywhere on Earth.

The first two Starlink satellites were launched in February 2018, and while SpaceX is reluctant to release concrete details, it is possible that internet operations will begin in 2020 and may cover much of the world by 2024 if it sticks to its FCC-imposed deadline.

Who is behind all this?


While SpaceX and their Starlink project make the news most regularly, there are actually several major players competing in this space. Here are some of the biggest:

  • Starlink: part of SpaceX, most high-profile. Possible service launch year: 2021.
  • OneWeb: low-profile, but well-organized and a major player as evidenced by its backing by investors from Richard Branson to Coca-Cola. Possible service launch year: 2019.
  • Telesat Canada: a Canadian satellite company with a test satellite already in orbit. Expected service launch year: 2020.

Other competitors include Samsung, Iridium, O3b, Space Norway, and even the Chinese government, though details on whether/when these will provide consumer internet access are sketchy.

How it works: lower orbits and bigger constellations


As cool as it would be, a satellite constellation won’t add any new scorpions or lions to the night sky. In fact, you won’t be able to see it at all. It’s more like a big net, completely surrounding Earth with a network of satellites. This technology isn’t very new. You use a constellation whenever you use GPS, for example.

What is new is the sheer scale. Starlink plans to start with 800 and build up to 12,000 more, which would double the current amount of satellites in orbit and be able to hit every square inch of the Earth with high-speed internet.

Why so many? It comes down to orbit height: most current satellite systems are in Geostationary Earth Orbit (GEO), which means they are 35,786 km/22,236 mi above the earth’s surface. The new systems will be in Low-Earth Orbit (LEO), which will put them 160 km/99 mi – 2000 km/1,200 mi above the surface. Simply put, our current satellite internet is slower because the satellites are farther away; bringing them closer decreases the signal’s travel time.


But here’s the catch. Imagine a flashlight shining on a bowling ball: if the flashlight is very close, the light only shines on a small surface area. The farther you move the flashlight, the more of the ball you can cover. Now replace the bowling ball with Earth, the flashlight with a satellite, and the light with data transmission. With the satellites/flashlights being closer to Earth, each one can cover less surface area, so you need a lot more to cover the same distance. On the plus side, most of the LEO projects estimate that their speeds will be comparable to current fiber optic capabilities, with barely any latency.

Enough with the tech stuff, when can I get it?

At this point, it’s less a question of if high-speed satellite internet will happen and more a question of when. If you live in the continental U.S, you’re in luck: most of the current projects, including Starlink, plan to start coverage there first, probably around 2020 or 2021. If there are delays, that data could be pushed back, but it’s probably safe to say that the network will be at least semi-global by 2025.

Details on how your new internet plan might look are scarce. With its huge coverage range, chances are it will be reasonably priced in order to attract a wide base of users, and just like any other satellite internet, you’ll have to get some equipment and a subscription to use it.

Why it’s great


  • Literally the whole world will have Internet. Rural areas are an obvious beneficiary, but even urban areas with poor connections or high-priced internet monopolies could benefit. It doesn’t take much imagination to see the myriad ways that a truly world wide web could change how things work.
  • Perhaps its biggest potential, though, doesn’t lie in giving already-connected countries a slight coverage boost. Starlink and OneWeb are both very focused on serving the underserved, putting a lot of emphasis on the potential for developing countries and isolated regions. 51% of the world currently has Internet access; these constellations could push that figure a lot closer to 100%.
  • And as if all that isn’t enough, if Musk wins the race, he plans to use the proceeds from Starlink to fund future travel to Mars, making humanity multiplanetary.

What could go wrong


Unfortunately, as with any complex project on this scale, there are many things that can go wrong:

  • It’s expensive. SpaceX has made great strides in manufacturing cheap satellites and has brought down the cost of launches by as much as 75%, but they’ve got a lot of satellites to launch, and unexpected cost overruns could still happen.
  • There are already some legal problems, especially surrounding who has the right to use the best frequency bands. There may also be some international issues — North Korea, for example, might not be too happy to have satellites hanging out above it giving out Internet left and right, and other countries will have regulatory standards to meet.
  • Then there’s space junk. It’s already a problem, and putting thousands of new satellites in orbit could make it much worse. SpaceX has already agreed to abide by much stricter standards than usual to ensure that all its satellites deorbit safely, but accidents happen. To combat this, SpaceX is investigating space-debris cleanup technology, and recently launched an experimental satellite called RemoveDEBRIS, which may help mitigate the problem.

In conclusion

Don’t cancel your Internet and start watching the skies just yet. The best-case scenario is still a few years away, and in that time a lot of things can happen in the fast-moving worlds of tech and space exploration. If you live in a developed country, this will likely just be a minor upgrade to your life anyway; the real potential for change is in developing countries, and the effects it could have there may spill over to benefit the whole world.

Image Credit: JCT600

Andrew Braun Andrew Braun

Andrew Braun is a lifelong tech enthusiast with a wide range of interests, including travel, economics, math, data analysis, fitness, and more. He is an advocate of cryptocurrencies and other decentralized technologies, and hopes to see new generations of innovation continue to outdo each other.


  1. More tech stuff! Are these bidirectional, and at what speed? What about gaps in coverage? What kind of equipment will be used? Where will the base stations be? What frequency bands will they use? Will we have to track the satellites? What happens to them when they reenter?

    1. Hey there! I actually wrote a really long reply to this with painstaking detail and a lot of sources, because I love talking about the tech stuff, but it seems to have gotten lost in the ether somewhere, so here’s a bullet-point summary of what I wrote:

      – Definitely bidirectional! Actually, omnidirectional–they will send and receive from the ground and will be optically connected to other satellites around them. Speeds are predicted to be somewhere between current cable speeds and current fiber optic speeds.

      – The only gaps would be country-sized; i.e, if North Korea threatens to shoot down any satellites over its territory, they likely won’t go there. Otherwise, even the penguins in Antarctica will have internet pretty much everywhere. A bigger issue is rain fade, and everyone’s been pretty quiet about that.

      – Equipment will be a ton of custom-made lightweight satellites, and satellite receivers with phased arrays to track LEO satellites. The receivers are rumored to be about the size of a pizza box and will go on houses, I believe!

      – Frequency bands will be parts of the Ku and Ka bands; a few different possibilities have been registered between 10 and 30 Ghz. The FCC filings are public and have all the info!

      – The launching company will be responsible for tracking and maintaining satellites, though I’d imagine that there will be lots of other entities doing it as well if it’s a worldwide scheme.

      – They should ideally burn up as they reenter, and SpaceX especially is going to a lot of trouble to make sure they do. If they don’t deorbit properly, they might become space junk though, which will require its own solution.

      My other comment was basically as long as this article and was a lot more thorough–sorry it didn’t get through for some reason! Hope the bullet points answered most of your questions :)

      1. “they might become space junk though, which will require its own solution.”
        If many more satellites are put into orbit, we will inadvertently wind up with something like a Dyson Sphere, most of it made of junk, around the Earth. Space is dangerous enough without all the nuts, bolts and other shrapnel whizzing around.

        If I had Elon Musk’s money, I would start a space salvage company that would recover space junk.

        1. He sort of is, actually–SpaceX is partnering with the University of Surrey’s “RemoveDebris” project, and they launched the prototype a few weeks ago:

Comments are closed.