5 ms·

Can someone gimme the gist of this, I'm not interested in reading for 39 minutes for an unknown point.

ghusbands 10y ago
Using satellites to provide good internet access is achieveable and avoids somewhat stagnant cable-based ISPs. Technical details and upcoming commercial ventures are discussed.
- mirimir 10y ago
  And with LEO, latency can be ~3 ms.
  - TheSpiceIsLife 10y ago
    Would internet require geosynchronous satellites? That's ~35,000 kilometers out rather than hundreds. That adds about 600ms.[1] 1. https://en.wikipedia.org/wiki/Satellite_Internet_access https://en.wikipedia.org/wiki/Satellite_Internet_access
    - mirimir 10y ago
      Nope, that's why Internet requires LEO, or at least MEO. It's basically a mixnet. Like how cellular works. Except it's not you that's moving.
      - TheSpiceIsLife 10y ago
        Objects in Low Earth Orbit have an orbital period of 88 minutes to 127 minutes. How many satellites would be required to provider continual coverage for any particular large city?
        mirimir 10y ago
        Given how it works, it doesn't make sense to focus on particular cities. Orbital footprints, perhaps, but that's probably not workable. From the article: Iridium used 77 satellites in polar orbits at 780 km. Planned LEO: LEOSAT 108 OneWeb 648 SPACEX 4000
        qmarchi 10y ago
        One of these things is not line the others.
        DiThi 10y ago
        SpaceX just launched and landed an orbital-class booster for the second time in history, a few days ago. It's not unfeasible to think they'll be able to launch their own satellites for very cheap, with the upcoming Falcon Heavy, with block 5 (improved version of the rocket for quick reuse) and after refining their inspection procedures.
        mirimir 10y ago
        Care to elaborate?
        sp332 10y ago
        They are each about a factor of 6 apart from each other.
        mirimir 10y ago
        I'm guessing that it's a power vs number tradeoff.
        xenadu02 10y ago
        Launching with SpaceX is already 50% cheaper than other providers, sans reusability. If SpaceX owns the network they can launch at cost. Let's say that puts them at a 70% advantage (When they perfect reusability maybe 90%). Satellite technology is also cheaper and smaller so you can put a lot more tiny satellites up with each launch. I have no problem believing SpaceX can blanket the globe with 4000 satellites at reasonable cost. As bandwidth needs increase they could simply launch more of them. For the sake of competition in the US (and access in other parts of the world) I hope they succeed. The only other risk I see (beyond losing money if the venture doesn't gain enough customers) is that nations like North Korea, China, Iran, etc will not be happy that their citizens can subscribe to an ISP they can't control or filter and without any way to block it. I can definitely see China (credibly) threatening to blow up the satellites if SpaceX sells access to Chinese customers.
  - neolefty 10y ago
    It also points out that you get actual light speed instead of fiber's 90% of light speed. I didn't realize there was a physical possibility of it being faster than ground. Mostly trivial compared to the other ramifications, but interesting nonetheless.

Using satellites to provide good internet access is achieveable and avoids somewhat stagnant cable-based ISPs. Technical details and upcoming commercial ventures are discussed.

mirimir 10y ago
And with LEO, latency can be ~3 ms.
- TheSpiceIsLife 10y ago
  Would internet require geosynchronous satellites? That's ~35,000 kilometers out rather than hundreds. That adds about 600ms.[1] 1. https://en.wikipedia.org/wiki/Satellite_Internet_access https://en.wikipedia.org/wiki/Satellite_Internet_access
  - mirimir 10y ago
    Nope, that's why Internet requires LEO, or at least MEO. It's basically a mixnet. Like how cellular works. Except it's not you that's moving.
    - TheSpiceIsLife 10y ago
      Objects in Low Earth Orbit have an orbital period of 88 minutes to 127 minutes. How many satellites would be required to provider continual coverage for any particular large city?
      - mirimir 10y ago
        Given how it works, it doesn't make sense to focus on particular cities. Orbital footprints, perhaps, but that's probably not workable. From the article: Iridium used 77 satellites in polar orbits at 780 km. Planned LEO: LEOSAT 108 OneWeb 648 SPACEX 4000
        qmarchi 10y ago
        One of these things is not line the others.
        DiThi 10y ago
        SpaceX just launched and landed an orbital-class booster for the second time in history, a few days ago. It's not unfeasible to think they'll be able to launch their own satellites for very cheap, with the upcoming Falcon Heavy, with block 5 (improved version of the rocket for quick reuse) and after refining their inspection procedures.
        mirimir 10y ago
        Care to elaborate?
        sp332 10y ago
        They are each about a factor of 6 apart from each other.
        mirimir 10y ago
        I'm guessing that it's a power vs number tradeoff.
        xenadu02 10y ago
        Launching with SpaceX is already 50% cheaper than other providers, sans reusability. If SpaceX owns the network they can launch at cost. Let's say that puts them at a 70% advantage (When they perfect reusability maybe 90%). Satellite technology is also cheaper and smaller so you can put a lot more tiny satellites up with each launch. I have no problem believing SpaceX can blanket the globe with 4000 satellites at reasonable cost. As bandwidth needs increase they could simply launch more of them. For the sake of competition in the US (and access in other parts of the world) I hope they succeed. The only other risk I see (beyond losing money if the venture doesn't gain enough customers) is that nations like North Korea, China, Iran, etc will not be happy that their citizens can subscribe to an ISP they can't control or filter and without any way to block it. I can definitely see China (credibly) threatening to blow up the satellites if SpaceX sells access to Chinese customers.
- neolefty 10y ago
  It also points out that you get actual light speed instead of fiber's 90% of light speed. I didn't realize there was a physical possibility of it being faster than ground. Mostly trivial compared to the other ramifications, but interesting nonetheless.