SpaceX Internet Satellite Network: Starlink

[mongo]

First I must say that I reelly do not know. But as long at SpaceX and/or Elon have the controlling majority in "Starlink company" I do not see that as a problem.

... but on the other hand: If Starlink find that they can get more profit/get the constellation going earlier by spreading out to different launch operators, would that really matter? The point is to get it going and generate profit to it's owners (SpaceX).



Yes, this is not the only way, and I don't even know if I would recommend it. I just pointed out that Starlink could be public while still have SpaceX private.

Sure, you can split the companies, agree with you there.
SpaceX is cheaper than anyone else and all profits go to SpaceX so I'm not seeing a case where it would make financial sense to use a third party (beyond renting launch sites or hitting the FCC timing limit, but that would require pre-ordering launcher capacity).

 

[Cosmacelf]

The majority of Internet bandwidth is consumed by video which is not very latency sensitive. Low latency packets, ie anythjng that isn’t video, can go via lowest latency links, while video can be relegated to higher altitude sats.

 

[electracity]

The majority of Internet bandwidth is consumed by video which is not very latency sensitive. Low latency packets, ie anythjng that isn’t video, can go via lowest latency links, while video can be relegated to higher altitude sats.

I can see the satellites serving different markets on different parts of their orbits:

1) high latency/low bandwidth: Existing sat phone and text business. Perhaps doesn't even always need a high satellite if the low satellites can go peer to peer.

2) high latency/high bandwidth: Netflix in the bush. The internet is a number of hops away, but there is substantial bandwidth.

3) low latency/high bandwidth: Today's cell service?

You're not understanding the order-of-magnitudes here. The low satellites are around 500km, while the higher ones are around 1400km (approximating, from memory). Berlin to Chicago, an approximation for average Europe-US distance, is about 7000km. So the high satellites are significant detours, adding about 40% to the distance, just about compensating for the speed-of-light savings. But Clarke Orbit satellites are 40,000km away. A single round trip to next door, via geosynchronous orbit, is more than 10 times as far as Europe-US.

It's not just the distance but the hops. A long path to a ground station is not going to be low latency.

 

[electracity]

Updated visualization:

 

[ecarfan]

That is a fascinating video. Interesting to note how, in the first phase, coverage only goes as far north in Europe as southern England and the Scandinavian countries are not served at all. Is that correct?

Later deployment phases will obviously cover those areas.

 

[electracity]

That is a fascinating video. Interesting to note how, in the first phase, coverage only goes as far north in Europe as southern England and the Scandinavian countries are not served at all. Is that correct?

Later deployment phases will obviously cover those areas.

I noticed that too. Probably too much complaining from Noway about Tesla service.

 

[Cosmacelf]

It still blows my mind that this sat network could halve the latency between far flung cities from current fiber.

 

[bxr140]

Sure, but if Starlink is seperate from SpaceX, does that impact the ability of Starlink to hire SpaceX for launches and what rate they are charged?

Thinking out loud, it could make it easier to hire SpaceX; in that arrangement of course Starlink would have to pay more than at-cost. If they are separate P&L centers that means SpaceX gets to take profit directly from Starlink right now as opposed to receiving equivalent monies flowed down through the annual operating plan of the higher level corporate entity. One could imagine that being a favorable relationship for both entities, at least from a launch perspective.

In reality such a corporate arrangement would probably manifest as a bulk buy (like there will be with all the constellations...), with a 'special rate just for you' for Starlink that's somewhere between the Gringo Price and SpaceX's cost.

Bigger picture, separating from SpaceX makes it easier for Starlink to influence the products from other launch providers. BO and all The New Guys are obviously quite a bit behind SpaceX; like all dynasties, nothing lasts forever.

Waxing a bit from the current topic, there's also the elephant of reality...its clear that it is REALLY hard to stand up an internet constellation. Oneweb's production shut down is evidence as is Elon's Redmond Tantrum, and in general everyone trying to play in the internet constellation game is finding it hard to lock down funding because, as it turns out, the only thing bankers care about is making money...

 

[electracity]

It still blows my mind that this sat network could halve the latency between far flung cities from current fiber.

Theoretical low latency under the condition of no queue on hardware that probably doesn't exists yet. Theoretically the boring company would be using hyperloop to run vehicles in a vacuum.

 

[Cosmacelf]

Theoretical low latency under the condition of no queue on hardware that probably doesn't exists yet. Theoretically the boring company would be using hyperloop to run vehicles in a vacuum.

Oh? Do you know what algorithm the video poster used to determine latency? Did he make a no queuing assumption?

 

[ZeApelido]

Spoke to my friend a bit at length about the work on Starlink (he is a wireless engineer at SpaceX). Keep in mind I have not paid attention to this thread so perhaps literally everything I say may just be repeating what's already been said

#1. Previous leads of the program got axed because - surprise! - they told Elon something couldn't be done in time. People in charge now are from a rocket engineering background, not wireless comms.

They closed Irvine office, then opened it back up because shockingly people didn't want to do a horrific commute to Hawthorne. Some concern that Seattle office might not stay open long term.

As the video above seems to illustrate, satellites will launch at 55 or 50 deg latitude (can't remember which) but orbit with different trajectories and oscillate between those latitudes N/S. So no coverage above 55 degrees latitude N/S.

Goal is to deploy "tens" of satellites early next year. Then hundreds by end of year. Best coverage will be at 50 degrees latitude. Worst coverage will be near the equator, as satellites are more spread out at the widest point. Equator will improve when a lot more satellites are deployed.

Satellites are like 3'x3' in size. Each trip they will dump them all out at the same time, then use internal controls to get them separated onto different trajectories.

Antenaes are like 18"x18". Interestingly this seems to be an area they have not spent much time on or optimized, not clear how compact it could get. First applications will be for big companies / governments then rural coverage. Made it sound like bandwidth would not be enough to compete with say Verizon in more dense areas.

He thinks definitely "possible" to have on cars, but not focus on that yet

Oh definitely latency will be much lower than other options. Big selling point.

That's all I can remember right now.

 

[BioSehnsucht]

Spoke to my friend a bit at length about the work on Starlink (he is a wireless engineer at SpaceX). Keep in mind I have not paid attention to this thread so perhaps literally everything I say may just be repeating what's already been said

#1. Previous leads of the program got axed because - surprise! - they told Elon something couldn't be done in time. People in charge now are from a rocket engineering background, not wireless comms.

They closed Irvine office, then opened it back up because shockingly people didn't want to do a horrific commute to Hawthorne. Some concern that Seattle office might not stay open long term.

As the video above seems to illustrate, satellites will launch at 55 or 50 deg latitude (can't remember which) but orbit with different trajectories and oscillate between those latitudes N/S. So no coverage above 55 degrees latitude N/S.

Goal is to deploy "tens" of satellites early next year. Then hundreds by end of year. Best coverage will be at 50 degrees latitude. Worst coverage will be near the equator, as satellites are more spread out at the widest point. Equator will improve when a lot more satellites are deployed.

Satellites are like 3'x3' in size. Each trip they will dump them all out at the same time, then use internal controls to get them separated onto different trajectories.

Antenaes are like 18"x18". Interestingly this seems to be an area they have not spent much time on or optimized, not clear how compact it could get. First applications will be for big companies / governments then rural coverage. Made it sound like bandwidth would not be enough to compete with say Verizon in more dense areas.

He thinks definitely "possible" to have on cars, but not focus on that yet

Oh definitely latency will be much lower than other options. Big selling point.

That's all I can remember right now.

"Tens" of satellites sounds like one or two launches, unless they just start piggy backing them on a bunch of unrelated customer launches 2 at a time. Which isn't unreasonable, "tens" could be enough to put up a whole single string of satellites for one orbit (or two, so you can test links between orbits that cross each other) and test that everything works as expected, then start back filling in the missing orbits.

I'm going to assume that 3'x3' is closer to 1m x 1m for sanity, but in either case I'm guessing that's the deployed form factor and that things will fold out from there. The antennas being 18" is interesting, I'm not sure we had much solid information or even rumor on the size of them. Not having the bandwidth to compete with (sub)urban internet is expected, even with the whole constellation with VLEO+LEO satellites in place, there's only so much coverage you can provide to a given area, and population dense areas will exceed that easily. They would likely still serve some commercial or public use (possible cellular backhaul, providing high speed internet to schools/colleges/etc) but they won't be directly competing with Verizon or such in the consumer space in those areas. Out in the sticks though I would expect them to sell service to individuals either directly or through a partner (so they don't have to deal with the customer support).

 

[ecarfan]

They would likely still serve some commercial or public use (possible cellular backhaul, providing high speed internet to schools/colleges/etc) but they won't be directly competing with Verizon or such in the consumer space in those areas

I don’t think Starlink is intended to compete with cellular data plans such as are offered by Verizon. It will compete with cable and phone company service to homes and businesses.

 

[BioSehnsucht]

I don’t think Starlink is intended to compete with cellular data plans such as are offered by Verizon. It will compete with cable and phone company service to homes and businesses.

No, they don't compete with cellular.

They will likely replace microwaves, fiber, and copper links for some cellular towers. An ISP's ISP if you will. Not everywhere, but you can bet on more remote cellular coverage becoming available once Starlink is active, because it makes backhaul easy.

They will certainly compete with the ISPs themselves in rural areas. In (sub)urban areas they'll mostly just compete with the business class services, as they are limited in how many connections they can provide per unit area (square mile/kilometer, take your pick)

 

[bxr140]

I'm going to assume that 3'x3' is closer to 1m x 1m for sanity, but in either case I'm guessing that's the deployed form factor and that things will fold out from there.

I'm pretty confident that the pathfinders are fairly good representations of the flight models, minus a few bits and pieces...and of course whatever improvements they've dreamt up over time. 3'x3' (or, 1m x 1m) coincides with the main spacecraft box, which is the white cuboid buried in the center of the pathfinders. The main support structure is strapped to the "top" of the box, and then all the bits and pieces are bolted to that structure.

The antennas being 18" is interesting, I'm not sure we had much solid information or even rumor on the size of them.

I think they're pretty much as we see on the pathfinders, again not considering any improvements since then. Its a pretty solid assumption that they're going to have 2 Tx and 2 Rx arrays (since that's what SpaceX has said). See the conversation upthread for our dissection of the pathfinders.

 

[Bobfitz1]

Fair enough. Though I would note that Elon has stated he wants to make Starlink sat production highly automated using mass assembly line techniques which will minimize humans and maximize robots, to drive down cost.
Now, we know that he wanted to do the same thing for Model 3 production, and that didn’t work out as planned. However, for Starlink to succeed it will require well over 10,000 sats in orbit, so if production can’t be highly automated the company is going to be in trouble.

Might someone more knowledgeable about Starling and Iridium speak briefly to how the type of services Starlink is designed to provide in massive amounts, overlaps or does not with the services Iridium will provide with their new constellation? I expect they have sized Iridium Next to provide 5 or 10X capacity over their current constellation. If successful, will Starlink eventually take over some business from Iridium, reducing the return from their several billion investment?

 

[bxr140]

Might someone more knowledgeable about Starling and Iridium speak briefly to how the type of services Starlink is designed to provide in massive amounts, overlaps or does not with the services Iridium will provide with their new constellation? I expect they have sized Iridium Next to provide 5 or 10X capacity over their current constellation. If successful, will Starlink eventually take over some business from Iridium, reducing the return from their several billion investment?

Yes, of course competition from Starlink (and Oneweb, and Telesat, etc.) will make it harder for Iridium, but that's always the case with business. There's always someone working on a better product than what you have to offer The new constellations are interesting for sure, as they will disrupt current products more in the way the original iPhone disrupted the mobile phone industry and less the way a new iPhone disrupts the current galaxy.

Most importantly, Iridium is available now, whereas a realistic timeline for starlink is on the order of many years. Iridium also has voice, ADSB (aircraft), and AIS (ships) in addition to Ku data, whereas starlink's product seems to just be Ku band data (happy for a fact check on that), so it's unlikely that starlink will wipe out all of Iridium's value. Iridium will also have a very established customer base by the time Starlink is ready for prime time, and all of those customers will probably have infrastructure--not to mention contracts--that keep Iridium in business.

Certainly copious global access to pure data [from a starlink-like constellation] will ultimately render things like sat-phones and ADSB and AIS obsolete, but especially with the latter two, that's not going to happen for a long time, and certainly Iridium is already working on developing their future roadmap, which could be up to an including being absorbed by one of the upcoming constellations. Interestingly, as technology evolves the mega sat constellations are probably going to come back under fire from geostationary satellites which can, from a practical perspective, be significantly larger and more powerful than a LEO constellation. Latency will always be an issue, but with the upcoming generation of geo pumping hundreds of Gbps all day and all night from a single asset and the next generation going into tera range, it is not unreasonable to imagine a scenario where LEO constellations and geocomms coexist.

Final thought: Note that Iridium has plenty of current competition (O3B, Globalstar, and Orbcomm, to varying degrees) so its not like they have a straight monopoly right now.

 

[mongo]

Might someone more knowledgeable about Starling and Iridium speak briefly to how the type of services Starlink is designed to provide in massive amounts, overlaps or does not with the services Iridium will provide with their new constellation? I expect they have sized Iridium Next to provide 5 or 10X capacity over their current constellation. If successful, will Starlink eventually take over some business from Iridium, reducing the return from their several billion investment?

In addition to all of @bxr140 's fine points, Iridium covers the entire earth pole to pole. Starlink is concentrated in the 55 degree and lower latitudes.

 

[Cosmacelf]

In addition to all of @bxr140 's fine points, Iridium covers the entire earth pole to pole. Starlink is concentrated in the 55 degree and lower latitudes.

If nothing else, sounds like Iridium will continue to have the Antarctic market sewn up

 

[BioSehnsucht]

Additionally, what Iridium loses in bandwidth and I assume latency they make up for not only in pole to pole coverage but the size and likely power requirements of the user terminals. Starlink is expecting pizza box sized user terminals, whereas Iridium terminals can fit in your hand. While undoubtably Starlink will steal some of the "higher end" of Iridium's business (i.e., anyone using it on a boat, commercial aircraft, etc will likely trade the size and power requirements for faster service, unless they need pole to pole coverage), Starlink absolutely cannot address Iridium's entire market, and Starlink might actually help sell new Iridium users (it's likely to become wildly known which will introduce new prospective customers who may find out it's a poor fit then learn of Iridium and go there). I wouldn't bet on this effect making up for the loss of users, but it might - the situations (boat, plane, etc) that most benefit from Starlink vs Iridium would already have found Iridium's speed to be a hinderance, and those that don't benefit from the speed would likely find the size advantage of Iridium to be preferable.

I don't see any reason why Starlink, Iridium, and Orbcomm can't coexist (as Iridium and Orbcomm have some feature overlap but again Orbcomm is useful for things that Iridium isn't the best choice for too). I actually would not be surprised to eventually see some accusations of a "Satellite mafia" (in the same vein as "Paypal mafia") since all three of these will likely depend heavily on SpaceX for launching and could potentially own the market between them.

Though Iridium and Orbcomm will likely continue to go with the most cost effective sufficiently reliable launch provider, so I could see Blue Origin undercutting SpaceX to gain market share in the future and thus these launches in the future being split among them. If ULA gets it's head in the game maybe they get a few too but their years seem numbered - at the going rate I expect that Boeing/Lockheed's ULA launch services will be allowed to die once Blue Origin is providing some form of competition for SpaceX and complete launcher redundancy (if something happens that prevents SpaceX launching, BO can keep going). If it wasn't for US DoD ULA wouldn't have even survived as long as it has, but DoD needed access to space, so it made sure the doors stayed open and there was at least two entirely different launchers (Atlas V and Delta IV) available. Orbital ATK (now Northrop Grumman Innovation Systems, but that's boring) might stay in the game via DoD assistance if necessary since they're the only solid fuel game in town really, and variety is the spice of life. But if I'm anyone other than Orbital ATK, SpaceX, or Blue Origin, I'm sweating my future survival in 10 years.

If I'm a satellite internet operator outside of Iridium and Orbcomm I'd be concerned, as Starlink will most likely eat their lunch. This will hurt both geosynchronous satellite operators (ViaSat and EchoStar for example who are both SpaceX customers) who can't win on bandwidth, latency, packaging, or power needs really... pretty much all metrics they lose. This will also hurt the other internet constellations being proposed / developed / etc such as OneWeb.

The market might be big enough for OneWeb plus the rest, but OneWeb will likely be left with serving customers Starlink can't (either an area is oversubscribed or OneWeb provides better coverage - I think they are intending pole to pole), or customers who want redundancy, or because they have skin in the game / politics (Airbus seems to be building OneWeb's satellites, so maybe EU governments will prefer it similar to preferring Arianespace over SpaceX). Another option is that OneWeb competes on price, as there will likely be room for them to do so as Starlink likely will be priced to what the market can bear in any given region (they'll likely undercut some of the incumbents but OneWeb might have room to undercut them). This would give you the option of paying less for less (OneWeb) or more for more (Starlink).

This is not all to say I want OneWeb or ULA or any of the rest to fail. I want a bright future where not only do they all succeed, but new ventures come forth and are able to find a niche or simply there's enough need for multiple providers in each marketplace. But as things are now, I expect some to fail, some to survive (perhaps with a pivot), ...