" the wheels are falling off of satellite TV"

[NYDutch]

SWAG TIME

Let's see 4,425 over five years. That's 885 per year and about 17 per week. Let us "assume" that they will put up 34 per shot. Maybe. And "assume" $100M per launch. That's what, about $13B just for launch? At least double (or triple) that for the costs of the satellites, and gosh knows how much more for the ground stations.

Let's be outlandishly charitable and "assume" the system can be built out for $25B. And essentially no revenue will come in until complete, or very nearly complete.

Let's cough and say they get a million customers. $25,000 per customer to recover. Even if it were a fifth of that, it seems unreasonable. Ten million customers? Well, maybe we're getting the total per revenue unit down. But it sure seems it'll be an EXPENSIVE service.


I would have to "assume" that since they have the go ahead to proceed, someone has figured it does not present a hazard.

SpaceX will be using their own reusable rockets, making the costs much less than conventional "throwaway" launch systems. And since SpaceX owns both the launch system and the satellites, I suspect their costs will also be considerably less than their competitor's costs. I imagine Musk is smart enough to place the initial orbits over the highest population areas first so revenue can be generated as soon as reliable coverage of those areas is in place along with the ground support systems. SpaceX is one of the most valuable privately held companies in the world, and they have a steady stream of launch customers now, averaging about one launch per month. If their new heavy launch rocket stays on track, by next year, they'll be delivering much larger payloads per launch.

 

[dare2be]

I imagine Musk is smart enough to place the initial orbits over the highest population areas first so revenue can be generated

Damn, another who doesn't understand. These are LEO (low earth orbits) like the space station. They can't be "positioned" over specific areas. They are not geostationary. They will move around the entire globe in a matter of hours.

 

[NYDutch]

Damn, another who doesn't understand. These are LEO (low earth orbits) like the space station. They can't be "positioned" over specific areas. They are not geostationary. They will move around the entire globe in a matter of hours.

Apparently I understand better than you do. LEO sats do not stay in a relatively fixed position like the geosynchronous sats that Dish and others use, but they can be placed in predictable, stable orbits where they pass over the same path with each rotation. Put enough sats spaced out in the same orbit and you can have continuous coverage of the area beneath that path. It's been done for many years now with spy and other sats.

 

[navychop]

Call me when his costs are 1% of his competitors.

And BTW, I "just happen" to know a bit about orbital dynamics. These aren't high apogee spy satellites. They will be polar orbits, slowly moving their footprint across the ground, with each and every satellite eventually covering the globe. No near circular orbit that keeps the same footprint each orbit.

 

[comfortably_numb]

Call me when his costs are 1% of his competitors.

This.

Unless he plans to solicit investment funding from a foreign socialist government (which is plausable), how will he get enough ROI to make this worthwhile and turn a profit? I’m not going to pay $150 a month for internet no matter how fast it is.

I’d like to see expansion of cellular wireless and small cells.

 

[Brussam]

The LEO satellites sound interesting but latency won't be just ground to satellite to ground to satellite to ground like Hughesnet. There will be a mesh network between the LEO satellites that will be part of the routing. More like ground to satellite to satellite to satellite to ground to satellite to satellite to satellite to ground.

After having been in a campground using a mesh networks to have many access points, the throughput takes a real hit in the traversing of the mesh.

 

[eddie willers]

This.

Unless he plans to solicit investment funding from a foreign socialist government (which is plausable), how will he get enough ROI to make this worthwhile and turn a profit? I’m not going to pay $150 a month for internet no matter how fast it is.

I’d like to see expansion of cellular wireless and small cells.

Not having to run a single wire to a single house...anywhere in the world. They catch up to AT&T's 100+ years in a single day.

First customer = billions.
Second to 100 millionth = pennies

 

[comfortably_numb]

Not having to run a single wire to a single house...anywhere in the world. They catch up to AT&T's 100+ years in a single day.

First customer = billions.
Second to 100 millionth = pennies

I don’t drink the Kool Aid that easily. We’ll see how it all pans out

 

[eddie willers]

The LEO satellites sound interesting but latency won't be just ground to satellite to ground to satellite to ground like Hughesnet. There will be a mesh network between the LEO satellites that will be part of the routing. More like ground to satellite to satellite to satellite to ground to satellite to satellite to satellite to ground.

After having been in a campground using a mesh networks to have many access points, the throughput takes a real hit in the traversing of the mesh.

Both SpaceX and OneWeb say the latency will be 25ms. I don't play videogames so I don't really know what that means, but the commenters on at ars technica seem to think that is pretty good.

 

[eddie willers]

I don’t drink the Kool Aid that easily. We’ll see how it all pans out

No doubt it's pie-in-the-sky right now, but with FCC licensing being approved, they have some serious plans. I need to learn more about the specific hardware involved and the spectrum that will be used (is Charlie sitting on a goldmine?) but Dish and DirecTV made a pretty penny with hard to install 18" dishes and no real advantage in dense populations. I am guessing that the LEO scheme means no hard to aim return path. If all that is needed is a small antenna (like an XM radio type) and only need to see "up" (no LOS problem) I would buy stock.

And I'm sure Bezos and Blue Origin are to be heard from. The estimates I have seen are that 10 billion is needed to start this. As strange as it is to say, that is, literally, pocket change to Bezos.

 

[NYDutch]

Call me when his costs are 1% of his competitors.

And BTW, I "just happen" to know a bit about orbital dynamics. These aren't high apogee spy satellites. They will be polar orbits, slowly moving their footprint across the ground, with each and every satellite eventually covering the globe. No near circular orbit that keeps the same footprint each orbit.

I'll bow to your superior orbital knowledge, but do look up the US KH-11 spy satellites... Hint: They're in LEO's. And Iridium currently covers the entire earth with voice and data communications with just 66 active LEO sats, but they do run in a higher LEO plane than the SpaceX sats will with significantly higher latency.

 

[NYDutch]

This.

Unless he plans to solicit investment funding from a foreign socialist government (which is plausable), how will he get enough ROI to make this worthwhile and turn a profit? I’m not going to pay $150 a month for internet no matter how fast it is.

I’d like to see expansion of cellular wireless and small cells.

The world wide Internet penetration rate is only about 51.7% as of June 30th according to the Miniwatts Marketing Group. That leaves a pretty big segment of the population remaining to be tapped, not counting those that would move from their current provider if the promised lower costs, higher speeds, comes to fruition. I have no idea who will come out on top in this race, but Musk and Bezos are both committing a lot of resources to LEO Internet, and neither one is known for taking up lost causes.

Internet World Stats

 

[dare2be]

And Iridium currently covers the entire earth with voice and data communications with just 66 active LEO sats

Communication satellites are a completely different animal and not at all a comparison to the Musk plan. A comm sat just has to be high enough and not too far away from other sats to be able to have a direct line of sight with other sats to avoid the curvature of the earth getting in the way. Latency is a big problem. Have you seen "live via satellite" news reports where the field reporter has to wait 3-4 seconds to hear the home reporter's question? That's mostly due to latency. That's why thousands of sats are needed just to cover enough of a particular footprint to be useful. Which means by definition they have to cover the entire earth at the same density.

That leaves a pretty big segment of the population remaining to be tapped,

I didn't consider this either, but it also leaves a pretty big segment of the population (think China and North Korea) that will have the opportunity to attempt (and maybe succeed) in hacking into it. No thanks.

 

[NYDutch]

Communication satellites are a completely different animal and not at all a comparison to the Musk plan. A comm sat just has to be high enough and not too far away from other sats to be able to have a direct line of sight with other sats to avoid the curvature of the earth getting in the way. Latency is a big problem. Have you seen "live via satellite" news reports where the field reporter has to wait 3-4 seconds to hear the home reporter's question? That's mostly due to latency. That's why thousands of sats are needed just to cover enough of a particular footprint to be useful. Which means by definition they have to cover the entire earth at the same density.

As I said, Iridium's sats are in higher plane LEO's than Musk, Bezo's, etc. will be using, which of course means they can cover the earth with fewer units, albeit with the higher latency we both mentioned, as well as needing higher power. "Internet" satellites are just another form of "communication" satellites of course, since they all use ordinary radio waves to transmit and receive their signals. In this case, all of them use or will be using hand off technologies from sat to sat similar to the hand offs used for tower to tower cell phone service as we move around. With the obvious difference that both the "tower" and the ground station could be in motion at the same time.

I didn't consider this either, but it also leaves a pretty big segment of the population (think China and North Korea) that will have the opportunity to attempt (and maybe succeed) in hacking into it. No thanks.

China, North Korea, and others are regularly hacking, or attempting to hack, into the current Internet provider system right now, so there's no change there. And the LEO systems will still be connecting to the existing backbone networks, so it's a given that attempts will be made on them as well.

 

[TheKrell]

Which means by definition they have to cover the entire earth at the same density.

Ah... I think due to orbital dynamics that if they want a certain coverage density at the equator, they'll have way too much coverage at the poles. I'm thinking of a polar orbit. I imagine the orbital inclination will be chosen such that they can serve the most customers where most customers are, and let the polls go uncovered and the equatorial region covered poorly. In other words, it won't be a polar (90deg) orbit.

 

[dare2be]

Actually, if not a polar orbit, and if they would want most coverage at 25-50 latitude to cover the mainland US, most of the orbits would have to be at a declination from an equatorial orbit, which would mean most of them would still spend equal amounts of time north and south of the equator. Which would actually mean that the equator would get heavy coverage.

 

[NYDutch]

Actually, if not a polar orbit, and if they would want most coverage at 25-50 latitude to cover the mainland US, most of the orbits would have to be at a declination from an equatorial orbit, which would mean most of them would still spend equal amounts of time north and south of the equator. Which would actually mean that the equator would get heavy coverage.

That could actually work fairly well in the earlier stages of implementation since southern Africa is one of the least penetrated areas of the world.

 

[TheKrell]

Actually, if not a polar orbit, and if they would want most coverage at 25-50 latitude to cover the mainland US, most of the orbits would have to be at a declination from an equatorial orbit, which would mean most of them would still spend equal amounts of time north and south of the equator. Which would actually mean that the equator would get heavy coverage.

Equal time north and south is true. The equator would still not get heavy coverage. Just picking an arbitrary ground track of a satellite at 55 deg orbital inclination results in more coverage nearest 55 deg north lattitude, and less for equatorial areas.

 

[sam_gordon]

Have you seen "live via satellite" news reports where the field reporter has to wait 3-4 seconds to hear the home reporter's question? That's mostly due to latency.

Actually, a dedicated video satellite hop delay (latency) is only 1/4 second. If there's multiple hops (from the Middle East to US will take 2-3 hops), multiply as needed.
Also, the delay has nothing to do with the field reporter hearing the question, but how long it takes for the video to get from the remote studios to the location.
MANY broadcasters are using cellular broadcasting (basically video streaming over the internet) technology now. That encoding and decoding delay is anywhere from 1.5 seconds to 20 seconds, depending on the quality of the bandwidth (longer delay = better quality).

I had the (un)fortunate experience of working with a LEO satellite phone service 10+ years ago. I'm not sure how many satellites they had, but we constantly had problems with the signal dropping out. Voice communications would use very little data. Getting significant bandwidth via LEO? I'm sure it can be technically done, but will it be cost efficient? Presumably the companies investing in the technology have figured that part out.

 

[rvvaquero]

My brain is like spaghetti, trying to read these past few pages. Do LEO's have wheels also? Are they coming off too?