Hacker News Reader

Comment by xenadu02 | Hacker News Reader

Having this kind of network is a good thing but the idea that any kind of satellite or wireless can replace wires is just fantasy.

Delivering just 10Mbps to 100 million customers is over 900 Tbps. There is absolutely no way a satellite network can clear that kind of downlink bandwidth. And that's just a tiny fraction of the market.

Physical cables/fiber represent several orders of magnitude more capacity and always will. So satellite is great for remote areas or unique situations but absolutely not a substitute for running cables.

View on HN

Your math assumes all 100 million customers will be downloading at 10Mbps at the same time, all the time. Obviously that is not the case.

Not "all the time", but at any one time - at the same time (eg: watching superbowl - but of course any kind of multicast/broadcast is a little different).

At any rate, it's mostly about the bandwidth each satellite has to route. I don't know what the likely number of subscribers for any one LEO sattelite will be? Will it really be on the order of 100 million notalso served by 4g cellular, dsl and other broadband?

Say for new York state its a large area and population total but the area where you might want/prefer satellite service isn't very densely populated?

Maybe there could be a "broadcast" mode where starlink users can "tune in" to a fixed number of "channels" which broadcast extremely popular content.

Couldn't the satellites themselves cache content that are for world wide events like this?

Edit: Even further, if they're launching 12000 of these couldn't that be a CDN network itself?

No, because HTTPS caching is not possible.

What I hear you saying is, they'd need to build out an actual CDN service that individual companies would have to sign up and use. This service would allow Starlink to be able to use their own certs themselves or have the owner's certificate.

My understanding is Cloudflare's service works this way. I believe this would be a good partnership between both companies.

I'm guessing the other approach is similar to Netflix's open connect. Where Netflix is given the option to run what they normally run on their open connect boxes but in a virtualized environment on the satellites themselves.

Now, I know that wouldn't seem all that practical if the service was seen as something anyone could sign up and buy because it isn't like these satellites are the same as EC2 instances and it isn't easy to just rack a new one but my gut feeling is that a very tiny portion of content is a substantial amount of bandwidth.

Now that would be interesting - 42.000 satellites running a cloud computing platform in space....

    There is absolutely no way a satellite network can clear that kind of downlink bandwidth.

Evergreen comment. Encoding and compression software and hardware continue to improve. The only constant in technology is change. Will we be able to do this sort of bandwith to LEO eventually? Yes. Can we do it this very moment? No one has tried.

There are underlying physical limitations to spectral efficiency. The best modulation schemes in use today can hit about 60 bits per second per hz of bandwidth. Meanwhile, single fibers top out above 40 terabit. To duplicate the bandwidth of just a single fiber link with a radio downlink, you'd need over 600 ghz of available spectrum. That'd require advances pushing deep into the thz gap, but beyond that, the physics of propagation are so different between 6 ghz, 60 ghz, and 600 ghz that it's hard to conceive of a single system covering so much spectrum. And that's not even getting into issues with geometric limitations of radio beam widths.

While change is constant, this isn't an argument that any change you can imagine is physically realizable.

Fiber is here to stay for a very long time precisely because it starts out with such massive fundamental physical advantages. It's also a mistake to assume innovation only benefits one alternative. We're still learning how to make and run fiber cheaper too.

> 60 bits per second per hz of bandwidth

Between one transmitter and receiver. Starlink is designed to cheat Shannon by utilizing SDMA on a truly massive scale. Every antenna in the system, both on the satellites and on the ground stations, is a large phased-array antenna, capable of simultaneously emitting different signals on the same wavelength in different directions, and capable of discriminating between multiple different signals on the same wavelength that come from different directions.

This is how Starlink intends to push so much throughput over just ~50 GHz of bandwidth. Not by pushing more bits per hz per link, but by filling the sky with hundreds of simultaneously visible satellites, and having each of them to all use all of the available spectrum at the same time on every link between any of them and any of the ground stations.

Phased arrays aren't magic. The directivity of the beam scales logarithmically with array size / element count. The limitations are such that you should think of it as being able to send Nebraska one signal while you send Kansas a different one. You can't get much more specific than that.

Elon has been very explicit about starlink never being competitive vs telcos in urban areas.

And independent people have done the math too. The system geometry maxes out at less than a single fiber: http://www.mit.edu/~portillo/files/Comparison-LEO-IAC-2018-s...

Phased arrays are not magic, but they are better than you make them sound like. When the directivity of the beam scales logarithmically with the element count, it just means you need a whole lot of elements.

The beam sizes are in the FCC filings, and they will be able to fully re-use the frequencies for ground stations that are >20km apart. (Not with the same satellite, you want adjacent beam spots on the ground to come from different satellites, but with literally hundreds above you at any time, that's not a problem.) Availability of bandwidth does not really limit them until the satellite counts are up to really hilarious numbers.

The aggregate throughput numbers in the paper you linked are correct, for the initial FCC filing. Since then they have multiplied the size of the constellation by ~7 and the throughput per satellite by ~3 (for 2.0 satellites). (ISL's are still missing, though.)

For what it's worth, satellite is typically around 1-4 bits per second per Hz.

Context is important, though. Network performance advances are always relative. If satellite fully-congested download speed improves 10x while hardwire link performance improves 10x, satellite will still be perceived as "too slow" by some segment. Our experience thus far has been that advances in applications tend to consume the bandwidth available. Whether that trend continues remains to be seen. (To be fair, I have never saturated my bidirectional 1GBps fiber link and I don't foresee ever doing so except for very brief periods of time.)

> Encoding and compression software and hardware continue to improve.

But these would apply to both wired and wireless, no?

If encoding tech is specific to wireless then not really , just like improving cable medium , number of cable pairs etc will not improve wireless , better transmitter or receiver will not impact wired

What encoding technology is specific to wireless? (Hint: none)

You are thinking just software , Hardware tech is very different for wireless. The benefits from beam forming, MIMO or any of the advancements in 3G/4G/5G only benefit wireless. There has been similar technical improvements in wired which only benefit that medium, they keep cramming more signal on the same size cable every generation.

Even in software at L3-7 the encoding tech is probably same of course, however protocols L0-2 is very medium dependent. The loss , noise, signal strength patterns all change depending on the medium so does

Even at L4 and above UDP based stacks benefit in wireless compared TCP based transport, if your application can handle it of course. The next gen HTTP is getting designed around UDP in some part to reflect the changing way people access the internet , a lot more traffic is mobile and wireless today so designing around wireless makes sense.

Personally even as app developer who works only with L7(HTTP) and L6 (SDP) I have specifically developed non trivial optimizations, workflows and APIs for mobile clients which were not needed or used by web clients due to different connection parameters of wired/ WiFi as opposed to wireless 2G/3G .

So yes there is lot of difference and most people working at any layer cannot ignore the medium.

I'm very aware of MIMO and the other physical layer technologies. But the original topic was about how wireless could somehow progress faster in these metrics than physical mediums. Mediums like fiber have progress much more in the last decade by adding more wavelengths and different modulation technologies compared to space technologies. SpaceX is only using MIMO in the sense that they have a phased array antenna, but that's not the same as your typical wireless router at home with multiple antennas. The other issue is space is severely power limited, so they are not going to get really high modulation schemes that fiber or cable have enjoyed (256+QAM) for many years.

> Encoding and compression software

What does that have to do with bandwidth ?

Sending less data over starlink (which is not tcp/ip btw).

I’d you can make 10Mb suddenly only require 900k of actual raw bandwidth by say adding a fpga or something then you can increase throughout to the users.

Technology only improves compression ratios to a certain point, and we passed that point with lossless compression decades ago. We're approaching that point with lossy compression now.

Going beyond that point requires the laws of physics (more precisely, the laws of information theory) to change, and that's not going to happen.

Hopefully in 2020 most internet traffic is encrypted. If encrypted traffic is compressible you're doing it wrong.

It would help if it was compressed pre-encryption, by the server.

If that big compression rates where achievable why wouldn't we be already using them? They would be standard in any network.

And even in remote areas, building out VDSL2 networks with relatively decent speeds (200Mbps down and 100Mbps up[1]) would be better, honestly.

Most homes already have a phone line and can support a DSL network... and in some countries a phone line is mandated by law, so the infrastructure already exists.

But, DSL is not nearly as "sexy" as thousands of satellites...

[1] https://en.wikipedia.org/wiki/VDSL#:~:text=VDSL%20offers%20s....

Even if DSL can get there, reliability of the connection in rural areas is often very suspect. A tree falls over in a winter storm, a bad crash that takes out an exchange box, a pole just plain falls over due to shifting ground, or maybe construction takes out a main trunk... these problems happen in cities too, but in a city you have maybe 1 km of infrastructure to worry about. In rural areas you could have 100 or 200 km of infrastructure just waiting for a calamity. Even just as a reliable alternative Starlink would be a big boon to rural users.

> Even just as a reliable alternative Starlink would be a big boon to rural users

Starlink is susceptible to cloud cover, rainstorms, lightning and other things that would interfere with signal quality (or render it unusable for periods of time).

Seems buried lines (either POTS for DSL, Coax for Cable, or Fiber) is the best solution. We just need to figure out how to incentivize it being built.

As someone who lived 50 minutes inland in Florida, no way.

Our DSL and cable options were limited, Comcast said they offered cable, but when they got to the line near the house, the signal was so poor they were unable to make it work without 50k in labor.

Then ATT came with DSL, the connection has routine outages for hours at a time. There is a technician servicing the DSLAM pretty much every week, there is literally no way ATT makes money where I used to live. Everyone owns 20+ acres, there's basically 10 homes in 1 square mile. And then the Hurricanes... Every 4 years all of the lines get blown down and it takes ATT 1-2 months to get it back online... There is way too many advantages for satellite in rural areas.

If I had to guess, there was over 200 outages (anything over 10 minutes IMO is an outage) in the 6 years I lived there.

GPON done properly is much cheaper to operate than DSL or COAX, it's just longer cables, less equipment, less real state, less electricity consumption.

Digging cables is what's expensive. You can use telephone poles for fiber too, but it exposes it to more outages. You can secure the lines with steel cables but that ads cost, and if you are deploying to poles then you were looking for going cheap.

Sounds like you have experience with above-ground lines. I specifically said buried lines - which are mostly immune to environmental issues.

Given the cost of burying lines is mostly in the labor/right-of-way stuff, we should be incentivizing fiber all over the place.

Helps that pesky Huwawei/5G problem the US keeps complaining about as well.

I haven't personally deployed VDSL2, but I have deployed DSL and other data-over-twisted pair systems including recent G.fast, G.hn and MOCA. I strongly doubt that VDSL can achieve those kinds of speeds reliably at the distances and densities needed to provide this service to the areas that need it most. In rural areas the copper lines are too long and too old to achieve these speeds (but they're fine for voice/POTS, so there's no mandate to upgrade.) Twisted pair is also really susceptible to noise and crosstalk, so in more densely populated areas once you start doing DSL to a lot of customers whose cables are bundled together they start to interfere with each other and everybody's service gets bad. In both cases the problem is signal to noise ratio - long cables mean poor signal, density means high noise, and twisted pair copper just isn't designed to deal with either very well.

What about VDSL2 Vectoring (G.vector)? Seems like it mitigates most of the crosstalk problems - but I admit I'm not an expert on DSL deployments.

It seems VDSL2 can achieve maximum throughput at run lengths under 1000ft from the DSLAM, with maximum performance beginning to degrade after. That sounds sufficient for most of the people currently stuck with slow access right now in the US (more rural areas outside major cities, but not actually remote as-in nearest neighbor is 1 mile away).

Also... surely solving this problem is cheaper than sending thousands of satellites into orbit every few years as they decay.

> It seems VDSL2 can achieve maximum throughput at run lengths under 1000ft from the DSLAM, and maximum performance degrades after that over distance. That sounds sufficient for most of the people complaining of slow access right now in the US

I would guess that this isn't right, but honestly I don't have the data either. My rough guess would be that only around 5-15% of people have <1000' of copper to their DSLAM, and I from what I've experienced with other technologies the speeds degrade very quickly after the max length, and only even work up to the max length in absolutely ideal scenarios. What I've seen of current copper infrastructure in the U.S. is pretty bad - lots of hand twisted copper lines with the ends exposed and rusting out, for example - that kind of thing would add the equivalent of at least several hundred feet of additional cable, and is very difficult to fix en-masse.

Honestly my gut feeling is that you pretty much have to run new cables to something like 80% of the U.S. to get good Internet service everywhere, and obviously if you're running new cables you should just run fiber. Looking at the cost of that compared to launching satellites - and accounting for the additional benefits of the satellites, like connectivity in extremely remote places and on the ocean - I'm not sure the satellites don't come out ahead.

EDIT TO ADD: A comparison point on the 1000 feet number is that this is in the same ballpark as how close a 5G tower needs to be to get real 5G service to your home or your cell phone. Current 4G towers are nowhere near that dense, and it's going to take huge amounts of money to get them that dense, even in the most densely populated areas.

Part of the reason this type of network isn't already built, is likely most folks living in very remote places simply don't need or want 1Gbps internet (or even DSL) - or at least don't know they want it.

There are folks with dial-up connections still, unfortunately, and not always because they have no other option. The more technically-inclined people here on HN wanting to work remotely and stream multiple 4k videos at once - they aren't majority of, say, AT&T's customer base. AT&T won't build this out if they think 1 in 100 potential customers will actually upgrade.

This will change over time, as a generation of folks that grew up with technology age and move into more rural areas. Hopefully demand will drive these network upgrades over time, and make it make financial sense to run new fiber in rural areas.

This could even be expedited with some sort of National Infrastructure Bill to fund it today, subsidized by tax payers since the benefits are ultimately so great.

I still think that's ultimately the better (and long term, cheaper) option verse thousands of satellites being launched every few years.

I don't think you're wrong here, but I would phrase it a little bit differently. I've worked in a lot of pretty rural areas, and from what I've seen in the last 10 years or so even folks in very rural areas want a lot of the same things from their service as techies do. Also, even places like the Bay Area with huge amounts of exactly the HN demographic the residential Internet infrastructure isn't exactly great and isn't improving very rapidly.

My take on the situation is that what most people want is more Internet for the same or less money. That means the only way to get infrastructure improvements is with competition. If AT&T did a bunch of upgrades and then gave their customers better service for the same price their customers would take it, but the customers aren't going to pay more than they're paying now and they're only going to leave if there's something better.

True, but their first customer is the U.S. Army which tells me this networks primary purpose is the same as the original intent of the internet itself. A mobile network that is available and reliable anywhere. Civilians usually get to do all the beta testing and provide feedback that improves the technology.

Also many of the places I have been looking at moving to don't even have POTS lines. This would fill the gap where POTS and VDSL is not available. I would certainly use VDSL if that were an option and use Starlink as a fallback in a very remote area.

> Also many of the places I have been looking at moving to don't even have POTS lines

Where are you, if you don't mind telling. I had thought, perhaps incorrectly, that in the US it was required by law to have POTS lines run to all homes - which led to it being installed during construction.

The land I am looking at is in the south western part of Wyoming. Could you link the law? My understanding is that once a line is run, there is a process to decom the line, but that no line is required for a home. Many homesteaders do not have phone lines or on-grid power. The only requirements I am aware of are around septic systems.

Interesting. I couldn't find the exact law to link to, but did turn up this WaPo article talking about 4 states starting to relax the requirement.

> The universal landline requirement has been repealed in Florida, North Carolina, Texas, and Wisconsin. There, new homeowners have no guarantee that they could order phone service at affordable rates, consumer advocates say.

https://www.washingtonpost.com/business/economy/landline-rul...

That seems more like a plan [1] to ensure that people can get an affordable line if they want it. I can't find any laws saying that all homes require a phone line. It also appears that recent discussions in the government were about changing money allocation from pots lines to internet specifically. I suppose we would need a land developer here to chime in.

[1] https://www.fcc.gov/general/universal-service

I've dug everywhere and can't turn up anything definitive.

The language, however, of "guaranteed access to an affordable line" does seem to imply a TELCO will have to run a line to your property at your request? Without fees (since that wouldn't be affordable)?

Perhaps, in practice, this resulted in all new construction having lines run since that was cheaper than doing it ad-hoc upon demand.

> I suppose we would need a land developer here to chime in.

Yes, hopefully someone out there knows more! Honestly curious now, since I had always assumed this was how it was.

There was a loophole in the various provisions created to ensure rural connectivity that allowed for companies to advantage of the fact that independent and competitive local exchange carriers were able to bill higher per minute fees than average. This led to a boom in free conference and long distance calling companies that would be spun up to terminate at one of these destinations, splitting the fee with the telco.

I mean, you're not wrong, VDSL in rural areas would be great! But it also hasn't happened.

The range just isn't there even for suburbs. The high speeds are only achieved with line length of less than 500m, and at 1600m it's similar to ADSL2... So not really a effective solution.

> And even in remote areas, building out VDSL2 networks with relatively decent speeds (200Mbps down and 100Mbps up[1]) would be better, honestly.

I live in the suburbs of Paris and the length of the phone line is still too long for VDSL2 (it's only available for distance less than 1 km) so I am not sure how feassible that is in rural areas.

I think they serve different purposes. Satellites serve areas where they can't run wires or it's insanely expensive for too few people. Think very rural areas and war torn areas. Imagine being able to get unblocked internet in China by using satellites? Or in Belarus. Or the middle of Amazon (rainforest)?

People keep bringing that up as a common use case, but how do you propose that people in China or wartorn countries will get their hands on the modems they would need to use Starlink?

Or even afford it?

People earning an equivalent of a couple USD a day, or nothing at all, aren't going to be able to afford to pay some US corporation for fast internet.

Fast internet probably isn't even a concern in wartorn countries, honestly.

Let's be honest - Starlink is almost exclusively going to be used by people living in rural areas within developed nations.

Implying the US government won't pay for the cost of an open internet where people can search for 'bad people square'

I'm not sure I understand your comment.

Why would the US Government (aka. Taxpayers) pay for internet use in other nations?

We already have expensive problems domestically that people would rather spend that coin on.

Winning 'hearts and minds'.

You drop access to the internet as whole in places where the internet is suppressed or non-existent and the populace learns just how bad they have things then they might revolt on their own.

I wouldn't say I 100% advocate for the idea but thats the thought process behind it.

That would generally be considered a violation of another country's airspace, and some countries would even treat that as an act of war.

You would also have to deal with the real possibility that those countries would simply shoot down any planes observed dropping Starlink base stations.

nonprofits air drop usbs full of news in north korea, violations happen all the time

US government airdrops?

There are a bunch or articles that talk about people using balloons to send material to North Korea.

The Human Rights Foundation even collects flash drives to load them up with content to smuggle into North Korea: https://flashdrivesforfreedom.org

For private groups, or individuals, sure you can get away with this.

For a government to do this? How could that not lead to an armed conflict?

Because if the government is the USA, who is really going to go to war with them over this?

You're not wrong, information/propaganda warfare is aggressive. Sometimes it makes more strategic sense then open warfare, and sometimes states can get away with it when they can't get away with open warfare.

How do governments airdrop supplies into war-torn countries today? I’ll admit I’m no expert but I see headlines all the time of neutral nations sending supplies to other countries who are fighting a war and that neutral nation isn’t getting dragged into the fight.

Governments don't airdrop supplies into war-torn countries today, excepting designated UN or MSF refugee camps, and they do so with the knowledge of one or both sides of the conflict so that the supply planes don't get shot down.

I would very much like to see a headline of a neutral nation airdropping supplies into a country engaged in war without getting dragged into the conflict. I'm not aware of that having occurred in the past 4 decades.

Do people not understand that satellite internet already exists today?

We don't drop satellite modems into war torn countries now, and we're not going to start doing that just because Elon Musk needs to check off one of his PR claims.

They're really expensive and big. Also, we actually do sometimes drop satellite internet into places.

Unless the pictures on Google are wrong, a Starlink base station is significantly larger than a satellite phone and is at least as large as a field base station.

And at $2000/pop, they're at least as expensive as the equipment we're currently not randomly dropping around the globe because of the expense.

Possession of unauthorized satellite equipment is as good as getting caught red-handed at a dead drop. A paranoid security state wouldn't blink at executing every such person as a spy.

Will Starlink even be able to legally sell internet services in China? Would they not have to comply with CCP's censorship policies and disconnect people at their will?

I strongly disagree with CCP's policies here, but that's not for us to decide.

I was thinking of the opposite use case -- an American traveling to China for work and getting uncensored internet.

But I mean once a Chinese citizen gets their hands on the equipment, I'm not sure how China would stop them. I suppose they could have roving bands of listening posts looking for the uplink signals.

I'm sure someone could figure out a way to take payments for them. Many rich Chinese citizens already have bank accounts outside of China.

Pretty sure an American company deliberately subverting another nation's laws to make a profit isn't going to be looked well upon, even if it is unjust CCP laws.

Regarding an American traveling to China - I think you're still beholden to the host country's laws, no? You can't break their local laws just because you're from out of town.

Not to mention the punitive actions CCP could take against Tesla (as they try to build out a business there) to put pressure on SpaceX to stop allowing Starlink access in China. That wouldn't be fair either, but Musk is CEO of both organizations and I'm not sure the CCP would distinguish them separately.

SpaceX isn't going to piss off China. The satellites are really just repeaters to the local base station. Starlink will be firewalled off from the world just like the rest of China (when the terminal is in China.)

>I suppose they could have roving bands of listening posts looking for the uplink signals.

I think this is pretty likely.

In the US the FCC has a whole set of systems designed to locate interference sources [1], and I've heard stories that they're extremely good at it. They even have a whole facility in Maryland dedicated to satellite monitoring!

[1] https://www.fcc.gov/over-air-spectrum-observation-capabiliti...

In rural areas, we have LTE-based FWA.

10 MHz wide channel with 256-QAM can push 100 Mbps down.

I'd argue if you get 100 Mbps down, you're not all that rural.

European rural :) It is a valley, tho, and if there wasn't that one tower, there would be 0 bps.

Economies of scale work massively in your favor in this situation. You definitely do not need to sell the bandwidth 1:1, and the more customers you aggregate the more you can oversell. With 1gbps upstream bandwidth you can comfortably provide 1000 customers with 10mbps Internet service and they will virtually never see speeds lower than 10mbps.

> Delivering just 10Mbps to 100 million customers is over 900 Tbps.

50 Gbps * 12000 satellites is 600 Tbps.

Most of those satellites will be over the middle of nowhere at any given time. But those customers don't use 10Mbps 24/7 either.

It also may not be limited to 12,000 satellites. They're seeking approval for up to 42,000.

Oct 2019: "SpaceX submits paperwork for 30,000 more Starlink satellites"

https://spacenews.com/spacex-submits-paperwork-for-30000-mor...

100Gbps microwave links have been shown on land. If you take that as a theoretical limit, 12k satellites can give you 1200Tbps.

That's not comparable to sending a signal from space.

I'm pretty sure it could with that DIDO technology that Rearden Labs produced. The think the sub company they started it under was Artemis. I always thought they're tech would work very well with satellite constellations.

Artemis was vaporware. It never made it into the field.

Vaporware is not an accurate description. I'm pretty sure you they have manufactured hardware that you can deploy for a test deployment. Also pretty sure they're working with the military for weaponization and wireless power transmission.

They were supposed to do a large rollout with sprint, and it never happened. The technology itself was vaporware; it only worked in their lab setting.

And yet such a satellite network is being built and deployed.

The people building this have obviously done the math and have a solution which they're implementing while you're busy saying that it can't be done :)

10Mbps to 100 million customers equating to 900 Tbps would be more akin to serving 1 billion customers, since on average someone isn't even close to saturating their bandwidth.

One satellite cannot route 900tbps of course , just like your local ISP cannot handle that either .

However over 40,000 satellites that’s just 22.5 gps per sat , quite achievable

Reply on news.ycombinator.com