The First Detailed Look at How Elon Musk's Space Internet Could Work (newscientist.com) 130
SpaceX has been granted permission by the Federal Communications Commission (FCC) to set up a vast network of thousands of low Earth orbit communication satellites. But the company has been tight-lipped about the project, known as Starlink. Mark Handley, University College London built a simulator based on public details from the FCC filings to understand the latency properties of the network. New Scientist reports: Although Musk has said he wants more than half of all internet traffic to go through Starlink -- Handley's simulation suggests that the project will be most appealing to high-frequency traders at big banks, who might be willing to fork out large sums for dedicated, faster connections. To create the simulation, Handley took what information he could from SpaceX's public FCC filings and combined this with his knowledge of computer networks. Initially, Starlink will consist of 4425 satellites orbiting between 1100 and 1300 kilometres up, a greater number of active satellites than are currently in orbit. There is only one way to arrange this many in a configuration that minimises collisions, says Handley. So he is confident that his simulation reflects what SpaceX is going for.
When sending an internet message via Starlink, a ground station will begin by using radio waves to talk to a satellite above it. Once in space, the message will be fired from satellite to satellite using lasers until it is above its destination. From there, it will be beamed down to the right ground station using radio waves again. Between distant places, this will allow messages to be sent about twice as fast as through the optical fibres on Earth that currently connect the internet, despite having to travel to space and back. This is because the speed of the signal in glass is slower than it is through space.
When sending an internet message via Starlink, a ground station will begin by using radio waves to talk to a satellite above it. Once in space, the message will be fired from satellite to satellite using lasers until it is above its destination. From there, it will be beamed down to the right ground station using radio waves again. Between distant places, this will allow messages to be sent about twice as fast as through the optical fibres on Earth that currently connect the internet, despite having to travel to space and back. This is because the speed of the signal in glass is slower than it is through space.
Re: (Score:2)
If you care enough about latency to have access to a massive grid of satellites as part of your route, then you're better off using a terrestrial radio link. You'll get faster speed and save tons of cash.
Re: (Score:2)
What about GPS? That's cheap and anybody can use it.
Re: (Score:2)
You RECEIVE a GPS signal. You don't upload to it. And it's not suited to doing much beyond broadcasting the current time. And they're about 12000 miles up. So no.
Re: (Score:2)
Very true, but just because they lack the ability to upload on the GPS network, does not mean a purpose built network wouldnt work. My point is GPS is available for all, and it is a "massive grid of satellites". Sure you will have to have a dish/antenna on the ground to be able to communicate and have full internet connection. But it can be done. And apparently they have some sort of idea on how to do it cheap enough to make it worth while. People also didn't think he would be able to reuse rockets, then pe
Re: (Score:2)
The distances among all the end-to-end points, be they game or gross margin, is a constant, and does not vary as does the speed of light in a medium, be it glass, copper, or over the air.
Of those, over the air is much faster.
Re: (Score:1)
The distances among all the end-to-end points, be they game or gross margin, is a constant, and does not vary as does the speed of light in a medium, be it glass, copper, or over the air.
Of those, over the air is much faster.
The speed of light though glass (fiber) vrs a vacuum is almost a rounding error, you don't see it until you get to 3 significant digits. I think the difference in distance is going to vary at about 2 significant digits, which tell me the surface fiber network will have the latency advantage.
Re:How about gamers (Score:4, Informative)
Glass is 31% slower, not a rounding error. Speed of light through air is much closer to matching your description. While this article from 2013 [extremetech.com] talks about using air-based conduits, I don't think it's reached full deployment yet.
Re:How about gamers (Score:4, Interesting)
Rule of thumb: light does:
* 1 m / 3 ns - signal in vacuum
* 1 m / 4 ns - signal in copper wire
* 1 m / 5 ns - signal in fiber optic cable
It's a lot slower in in a bent fiber optic cable, of course, but long runs are effectively straight.
Re: (Score:1)
The speed of light though glass (fiber) vrs a vacuum is almost a rounding error, you don't see it until you get to 3 significant digits.
How does refraction work then?
Re: (Score:2)
Given the choice of delivery of information OTA or struggling through a non-superconductive conduit, I'd like mine airborne, please.
Signal transmission geometry (Score:2)
The distance between end-points doesn't change, but the path between them does.
And going via satellite is potentially much lower than any other alternative, at least to get halfway around the world - after all, the speed of light in a straight line through vacuum is pretty much the limit, nothing else can be faster. And you can't send signals in a straight line on Earth - the Earth gets in the way. Realistically you've only got line-of-site of 80km or so - even at 10,000 feet altitude that only increases to
Apples to Oranges (Score:5, Interesting)
Problem is, fiber on the ground is faster as the distance is shorter.
Only true in general for relatively short trips with fixed destinations.
Unless you live in some remote place that depends on geo-stationary satellites for internet, you are better off going along the ground, at least where latency is concerned.
A) These are not geo-stationary satellites SpaceX is proposing. Geostationary orbit is about 35,700km away versus the 1200km being proposed here. That difference is very significant. B) Ground is only faster in some use cases but not all and the longer the transmission the less advantage it has.
Re: (Score:1)
Problem is, fiber on the ground is faster as the distance is shorter.
Only true in general for relatively short trips with fixed destinations.
Which is basically what this is. No matter how you slice this, the ground based path will be physically shorter and thus has a latency advantage, even with LEO satellites in the mix.
Unless you live in some remote place that depends on geo-stationary satellites for internet, you are better off going along the ground, at least where latency is concerned.
A) These are not geo-stationary satellites SpaceX is proposing. Geostationary orbit is about 35,700km away versus the 1200km being proposed here. That difference is very significant. B) Ground is only faster in some use cases but not all and the longer the transmission the less advantage it has.
I never labored under the illusion that we where discussing geo-stationary orbits. But as you point out, the distance adds huge amounts of latency.
As I see this, the LEO data route has to go 1200km twice to get next door, so you have to gain 2400km's worth of latency somehow over the ground based system. I don't think the v
Re: (Score:2, Informative)
I am a network designer, or network architect if you prefer. You are correct. A mesh satellite network which could beat a typical terrestrial link is not possible at this time. The only time Musk's current plan might compete on latency are for links across the pacific. Even then, often the terrestrial lines will win. LEO might be able to provide greater resistance to jitter, but there really is no way to predict that until typical traffic volumes, patters, and types are empirically provided as a test-load.
C
Latency (Score:2)
Which is basically what this is. No matter how you slice this, the ground based path will be physically shorter and thus has a latency advantage, even with LEO satellites in the mix.
A) Unless you are talking about some specific use case, ground based paths are NOT always physically shorter because they are not all point to point connections. B) Ground based routing of any significant distance routinely has to go through more devices and at slower speeds through fiber/copper. C) Latency is not just a function of distance
But my point here is that if you are looking at latency, the shortest route unusually wins and LEO orbits add quite a bit of distance.
Not when you are talking about distances the size of a continent. It's 4500km from NYC to LA across the earth's surface assuming a relatively straight line connection
Re: (Score:2)
As I see this, the LEO data route has to go 1200km twice to get next door, so you have to gain 2400km's worth of latency somehow over the ground based system.
First, why 2400 km? There doesn't seem to be a need to jump twice through the same satellite. Second, the Pythagoras theorem significantly diminishes the difference for medium distances.
Re: How about gamers (Score:2)
Re: (Score:2)
>Problem is, fiber on the ground is faster as the distance is shorter.
The distance is shorter, but the speed is much slower - typical fiber optic cable has a refractive index of 1.467, which means that light travels 46.7% faster through vacuum than cable.
Re: (Score:2)
Problem is, fiber on the ground is faster as the distance is shorter.
The ground distance must be shorter by at least 32% because the speed of light in fiber is 32% lower than the speed of light in free air.
Re: (Score:2)
Sorry, that's copper that's 30% slower. Some fiber connections are about 99,7% the speed, plus microwave though air is a rounding error from a vacuum.
So a series of microwave links or fiber still is shorter time wise.
Re: (Score:2)
Re:How about gamers (Score:5, Informative)
The lower latency could be useful for games, too.
Probably not, because the potential savings scale with distance. Halfway across the world and back is 40000 km/300000 km/s = 133 ms at light speed, at c/1.7 it's 226 ms so at most 93 ms to save. But I wouldn't try to play a twitch game at those ping rates while at reasonable gaming ping times the gains are mostly eaten up by the base latency. Unless you really desperately want to play with your guild from another continent, it's not going to matter much. It mainly matters for HFT where your buy/sell orders arrive a few milliseconds before the competition.
Elon Musk (Score:3, Insightful)
Solving the worlds problems, step by step.
Re:Elon Musk (Score:4, Insightful)
Or more like building the technological infrastructure needed for a Mars colony, step by step.
Dual purpose (Score:2)
Or more like building the technological infrastructure needed for a Mars colony, step by step.
Maybe but a lot of those problems are the same problems we have here on earth. We need electric powered vehicles to reduce oil dependence. We need low cost to orbit rockets. We need solar powered homes. We need cheaper/better tunnel making. We need more ubiquitous internet access globally. Whether you like Musk or not, you have to admit he's working on solving serious and important problems. (and if you don't think those are serious problems then you don't understand the problems) The fact that ther
Re:Elon Musk (Score:5, Insightful)
...and making them viable.
Re: (Score:2)
What is missing is the NSA collection point.
This runs into the same problem that started that whole thing - having a large portion of traffic running through one company (or country as was the case).
If one entity "handles" the traffic then they can abuse users by sniffing that traffic.
I'm not saying there's any good solution to get around that, just be aware. L3 and other entities handle large portions today.
Re: (Score:2)
Re: (Score:2)
Holey Fiber, [Star]man! (Score:5, Informative)
Just dropping by to mention that, while existing fiberoptic networks have index of refraction around 1.7 (so signal speed is c/1.7) , there is a relatively new thing referred to as "holey fiber." It's essentially analogous to microwave hollow guides, with the hole pattern sized to match the TE/TM modes of the injected light. The speed thru these waveguides is close to the vacuum limit.
Re: (Score:1)
None of it's in the ground, but would be a lot less polluting than hanging 4400+ sats in LEO. Go into any town in the US and look around at the pollution of cable, phone, electrical lines + the madness of cell towers multiplying in the night.
This isn't disruptive, it's another less-is-more problem. Latency is a problem only for the rich and those in need of killing orcs online. Once congestion builds, another 4400+ will go up. Then they'll use lasers for security... and a second skin will cover the earth, a
Re: (Score:2)
And while holey fiber provides improvement over classical fiber, it does not provide improvement to the speed of light OTA.
Re: (Score:2)
And so, if both transports are of the same velocity, then the shorter distance wins the speed contest.
Problem is: speed serves only a few individuals for the effort and money spent, until a new medium that's either shorter or more dense (!!!) wins. Fiber's rarely been a bad investment. However, radio is a finite resource, even with better modulation schemes.
Re: (Score:2)
And so, if both transports are of the same velocity ...
Ah, therein lies the rub.
Transports don't move. Light moves within the transport and no two transports (unless they are the same) allow light to travel at the same speeds.
Look at the speed of light in water, copper, aluminum, fiber, OTA ... and appreciate that "light" is not limited to the visible spectrum.
... until a new medium that's either shorter or more dense ...
Sorry. Wrong, and wrong. The medium carrying information, even if shorter, does not reduce the distance say, from New York to London. And, you went the wrong way regarding density. Light travels faster i
Re: (Score:2)
It's true, transports don't move. They're layer 1.
If both have the same velocity factor, then the speed of light is the same between the two. For satellites, there is a mixture from the point of terrestrial origination in velocity factor. Once in the sat, sat-to-sat is fixed, then there is the next velocity factor of the signal's return to earth. Full duplex communications across the full link should be about the same, with some modest but controlled jitter.
The proposed waveguide fiber has the characteristi
Re: (Score:2)
It's true, transports don't move. They're layer 1.
Sorry. TL;DR because bullshit.
Re: (Score:2)
...it does not provide improvement to the speed of light OTA.
Nor does it solve the two most important terrestrial problems that exist with today's Internet: high speed coverage in rural areas, and last mile delivery that bypasses the local wire monopolies. Starlink promises to solve both of them. I just wish it were available RIGHT NOW, so I could tell both AT&T and Mediacomm to piss off.
Re: (Score:2)
Well, we are getting close to Christmas (void where prohibited), so let's add municipal WiFi and fiber to the list.
Re: Holey Fiber, [Star]man! (Score:2)
Actually starlink has another advantage disadvantage. Weather. A massive storm hits you will lose starlink access in the short term like you would cell reception. But once it passes you have full access again. No rebuilding of towers or stringing new lines in.
I wouldn't want it as my only access but damn it would be a great backup. Depending how it does with roaming connections it might make an awesome cell provider.
Re: (Score:2)
Radio at those frequencies goes through clouds well. It doesn't have to penetrate at lower frequencies, clouds, fogs, and compete with bursts (lightning).
The problem with cellular is the same as with Imsat phones: latency. The speed of light only goes so fast.
Re: (Score:2)
I'm not so sure about the disadvantage. There's plenty of microwave/radio spectrum to which clouds are transparent, and the power from thunderstorm radio noise, while broad-spectrum, is heavily concentrated at lower frequencies.
Re: (Score:2)
Re: (Score:2)
Typical fiber gets you 0.7c. Not c/1.7 (.588c).
Re:Haha, this won't last. (Score:4, Insightful)
Iridium et al. Just a bunch of space junk now.
Is it? Iridium's problem was lack of customers who needed the service and had money to pay for it. Prices where so high that few people in areas that didn't have cellular service already could afford to pay the subscription fees. Their business model wasn't viable.
In this case, there is actually a different business model and possible paying customers who actually have money and a need for internet service. Are there enough of them who have enough money to make this viable? Maybe... Given Musk's access to exceptionally low cost launch rates, this might just work out this time. I'd not bet on it either way myself.
Re: (Score:2, Insightful)
Iridium's problem was...
Actually, AMONGST Iridium's problems was that it was not designed as a data transport system, and the horrible data capabilities that they offered were based on running a modem over their voice lines.
Iridium Next will expand those offerings to 128 kb/s to mobile, 1.5 Mb/s to marine and 8 Mb/s to fixed land stations.
This leaves a lot of room for improvement and Musk might force iridium's hand to provide faster upgrades than their 20 year turn around on iridium to Iridium Next
Re: (Score:2)
Now, these satellite services are great for under-served areas of the world. The problem is, the customers available, those willing and able to pay, and needing service in those under-served geographies, there aren't that many of them really. Most people in the under-served geographies are poor, or they don't need this service. But put all that aside for a moment and consider this. The majority of people worldwide live in (or near) cities and already have internet and cell and WiFi service.
The people paying the telephone and cable monopolies are under served also.
Re: (Score:2)
Patent: The internet ... but in space.
You aren't thinking big enough. Patent "space" - with an Internet. Then shut down anyone doing anything in space with data communications that go through more than one hop...
Shortwave Trading (Score:5, Informative)
https://sniperinmahwah.wordpress.com/2018/05/07/shortwave-trading-part-i-the-west-chicago-tower-mystery/
Re:Shortwave Trading (Score:5, Interesting)
Re: (Score:2)
Actually... The way this works now is you buy server space adjacent to the exchange and configure and run your trading platform software from there. That way your latency is as low as possible as your algorithms are running in the next room. Latency to the human controller doesn't really matter anyway as the trades are triggered by software anyway.
Surfaced based communication channels are soooo yesterday.
Re: Shortwave Trading (Score:3, Insightful)
For example: Gold price in New York has a relation with gold price in London. If you see Gold price diving in London. You may want to fast sell gold in New York before it dives too.
Who ever places this order 1st gets the best result
This cannot be done without distant trade links where latency matters.
Re: Shortwave Trading (Score:4, Insightful)
It's amazing the stupid things that drive technological progress.
Oh well. New tech is new tech.
Re: (Score:2)
This is not true. Most HFT is based on being a few meters closer to the exchange.
Re: (Score:2)
Re: (Score:2)
Re: Shortwave Trading (Score:1)
This technology will be equal for all and available for all.
Trading will be faster but no one will profit.
Re: (Score:2)
The problem with fast trading is that a bird 1100 km up will never be faster than a fiber 30km long on the ground.
That is not a selling point. Ever.
Re: (Score:2)
> a bird 1100 km up will never be faster than a fiber 30km long on the ground.
Not for 30km, but for 10's of thousands of kilometers, as is relevant to trading between international markets, yes it will. The key is that light travels about 47% faster in vacuum than in fiber-optic cables.
Re: (Score:2)
It's not the hop distance that's relevant to gaming and high-frequency trading though, but the full end-to-end latency
So, let's see where breakeven point is:
By land, great-circle effective distance = 1.5 *angle * 6400km
By 1000km altitude orbital great-circle = angle*7400km + 2*1000km
1.5 *angle * 6400km = angle*7400km + 2000km
angle*6400 = angle*4900 + 1300
angle*1500=1300
angle=0.87
equivalent great-circle distance = 6400*0.87 = 5,600km
So, for any link greater than about 5600km the satellite link will effective
Re: (Score:2)
Don't forget, you're aiming at a moving target. That diagonal transmission may end quick as the satellite goes under the horizon and another satellite needs to be located and aimed at. I assume the radio signal has to be directional for maximum bandwidth.
Re: (Score:2)
Well, you're going to be aiming at a moving target regardless, and the steeper the slow, the less the angular speed of the satellite as seen from the surface station. It's actually hardest to keep a tight focus on the satellite when it's directly above you, and gets easier the further away it gets.
You probably do want a directional signal though, both to minimize interference with other signals, and to minimize the power consumed. And it may well be that the satellites will all be aimed straight and only c
Re: (Score:2)
Yea, that sounds about right. Be interesting to see how it works in practice.
Re: (Score:2)
Light is always traveling at the speed of light.
The qualifier you're looking for is the medium.
And it is impossible for light to reach the theoretical limit of, "in a vacuum." There is no vacuum. That would violate Hiesenberg's Uncertainty Principle, for one thing. For another impediment, see "quantum vacuum."
Re: (Score:2)
"in a vacuum" is implied any time other qualifiers are left out. That's the only common reference point (and the inevitable virtual particles are *part of* that vacuum). Much like all gravitational potential energy is always measured as negative in space, because the energy at infinite distance is the only common reference that makes sense to use as zero.
Re: (Score:2)
The vacuum does however violate the second law of thermodynamics, which is why quantum foam has been postulated.
Re: (Score:2)
Nice use of buzzwords, but you could have increased the "cool points," if you'd thrown in, "blockchain."
The complete lack of anything at all would be, with certainty , a "zero point."
Armed with the new knowledge I have gifted you, go read (probably for the first time) Heisenberg's Uncertainty Principle.
Astounding (Score:1)
When sending an internet message via Starlink, a ground station will begin by using radio waves to talk to a satellite above it. Once in space, the message will be fired from satellite to satellite using lasers until it is above its destination.
Why it is almost as if they actually have an L2/L3 network! How could the reporter actually type the above text without passing out from the sheer excitement.
Re: (Score:2)
A better question would be: What's the cumulative latency of the bucket brigade?
Already out of date.. (Score:5, Informative)
https://www.theverge.com/2018/... [theverge.com]
https://forum.nasaspaceflight.... [nasaspaceflight.com]
Re: (Score:3)
SpaceX has actual data, as the two test satellites have been in roughly that orbit. Worst-case latency will be higher.
For reference, one way speed of light delay to geostationary is at least 120 ms.
the FCC? Authority? (Score:3)
Who gave them the authority to do that? They should probably only be allowed to have a say about the airspace above the US.
Re:the FCC? Authority? (Score:4, Informative)
The FCC only has authority over the U.S. radio spectrum. Now he's got some 200 more approvals to go.
Iridium did the same thing on a smaller scale back in the 90's when the company was owned by Motorola . The tech was worked out very quickly. It was the politics that slowed them to a crawl. Motorola found out exactly how hard it was to get all the world governments and incumbent telcos to agree to give them a sliver of spectrum. They had initially planned on direct satellite to phone communications until several telcos raised their hand and said no way, they had to go through ground stations where they could listen in on the conversations as well as charge access fees.
Musk just fired Starlink's executive team because they couldn't meet his timeline. If they were bogged down trying to get spectrum, it's no surprise - it took Motorola years and thousands of meetings and bribes to pull it off. Motorola's final approval only came after a Motorola lobbyist waited until a specific country's representatives fell asleep. The lobbyist knew the reps were going to vote against approving the worldwide spectrum at an international telecom conference. The lobbyist delayed until 4am in the morning when very few representatives were present and the known going-to-vote-no guys were present but asleep. Only then did he bring Motorola's request up for a vote. It passed and Iridium was born
After all that drama, Motorola/Iridium ended up with a very skinny slice of spectrum. More spectrum means more bandwidth. Bandwidth is something Iridium isn't known for.
Sharks with laser beams, yes! (Score:5, Funny)
I still prefer the notion of a network of ocean line-dancing sharks supporting a mesh net of laser beams just skimming the surface.
The training and fish bill is high, and this network facilitates phishing too, but you can't have everything.
Rgds
Damon
Re: (Score:2)
Yes, this product would be "SharkLink". I'll get my (shiny, buoyant) coat.
What people need are specifications (Score:5, Informative)
Fortunately, I have them.
Space Communications Protocol Specifications [wikipedia.org]
Consultative Committee for Space Data Systems [ccsds.org]
CCSDS Technical Specifications [ccsds.org]
Space Assigned Numbers Authority [sanaregistry.org]
Spacecraft ID list and manual [ccsds.org]
Disruption Tolerant Networking [nasa.gov]
Exploration and Space Communications at NASA [nasa.gov]
Free Space Optical Communication [wikipedia.org]
Satellite configurations (Score:1)
"There is only one way to arrange this many [satellites] in a configuration that minimises collisions, says Handley."
Hadley's statement above is actually false. It suffices to present two viable configurations to prove his statement false.
There are at least two basic configurations for plotting the vertices in a geodesic dome: class one and class two. Thus there is more than one way to arrange this many [satellites] in a configuration that minimises collisions.
If the satellites are not plotted like the v
It’ll be great until ... (Score:2)
67 comments in... (Score:2)
What do we know? About SpaceX SAT architecture, network topology and service niche opportunity?
Even the simulator doesn't layout the business case much less a technological opportunity for SAT use.
Re: (Score:2)
The architecture and protocols, I have listed above.
A range of simulators that could be helpful:
https://github.com/mlab-upenn/... [github.com]
http://ssfnet.org/homePage.htm... [ssfnet.org]
https://www.nsnam.org/ [nsnam.org]
Re: (Score:2)
What do we know? About SpaceX SAT architecture, network topology and service niche opportunity?
Well, this is slashdot. Full of uncountable self-appointed experts in all matters, both technical and otherwise. We still don't understand why business people all over the world don't flesh their concepts out with input from slashdot so as to make sure they are only wildly successful.
My personal opinion is that if SpaceX has $$ that they are willing to throw at this, they probably have done some of the basic back of the envelope calculations necessary to justify it. Unless they like lighting $100 bills
What? (Score:3)
Handley's simulation suggests that the project will be most appealing to high-frequency traders at big banks, who might be willing to fork out large sums for dedicated, faster connections.
Well that's just dumb. High frequency traders at big banks merely locate their data center / computing presence in close physical proximity to the point where the trades occur. Relying on a massive, expensive space network to come into existence just for high frequency trading is absurd.
This will most appeal to the millions of people that do not have broadband. The money to be made is in the masses, not in "high-frequency traders at big banks".
Re: (Score:2)
High frequency traders at big banks merely locate their data center / computing presence in close physical proximity to the point where the trades occur.
Not only in close physical proximity (eg Paternoster Square in London) but also for direct line-of-sight via microwave link. Eg the HFT microwave links in Aurora Illinois. [youtu.be] Those will always be faster than a round-trip via satellite.