Which launch company would they go with? SpaceX? Even then, they are splitting costs of the rocket (most expensive part) and fuel, while SpaceX themselves can just pay for fuel on an already flown booster. SpaceX are also paying cost, rather than retail prices.
Sure, there are a few more costs, the point being that SpaceX has a supreme cost advantage.
Maybe not. The receivers are not low-tech, they can ban possession and track anyone using it by the signal. The great firewall is not perfect either, it just needs to be good enough to discourage most of the population from even trying and tag everything else as suspicious activity.
I'm hopeful that people under repressive governments and censorship will be able to leverage satellite services for uncensored access.
With SpaceX's scale and low cost, I suppose it's even possible they could adopt a model where China tries to shoot down their satellites when they are over China and when they're successful SpaceX just replaces them with spares. It's possible that launching satellite-killer missiles will cost China significantly more than replacing satellites will cost SpaceX which might make it a non-viable approach for the Chinese. (Note: I don't know anything about anti-satellite missiles, so just idly speculating).
SpaceX really doesn't want to let too many of their satellites get shot "down". Shooting a satellite doesn't really make it go "down" immediately, instead it turns it into a debris field. At most of the altitudes SpaceX is planning on putting these satellites that debris would fall back to earth "quickly", but "quickly" is probably measured in months not hours. In the mean time that debris can crash into other sattelites. Shoot down enough satellites and you're going to get a chain reaction where the debris ends up taking out more satellites, generating more debris, and so on until the entire orbit is basically empty of functional satellites.
I doubt SpaceX is going to go to war with China, and if they did I doubt they would win. China, like other major powers, has anti satellite weapons.
Most likely beamforming will be used to avoid servicing Chinese territory as much as possible, and they will not try to have people smuggle receivers into China.
I'm curious why people aren't seeing China as, at least potentially, a very big customer. They have hundreds of millions of people with limited to no internet access, in spite of a rapidly increasing level of national wealth. Their government could change that, instantly, with a check.
Not that I'm in favor of a censored internet but it's a reality of China and it would perhaps work even better with something like the Starlink. Satellite internet relies on ground stations to know what to beam out on requests. China would operate their own regional ground stations resulting in a complete centralization of all access from China.
Only reason I see this not happening is if China chose to develop their own similar solution, which is perhaps a real possibility.
I don't see China as a potential customer because I don't see the CPC allowing any company not under their control to run China's telecommunications infrastructure. Especially not a company under US control.
It would be silly to think this will actually come to fruition - 30k satellites would represent a 3000% increase in the number of active satellites in LEO, which is already by far the most congested orbital altitude.
That said, it would be quite exciting to see all those launches, and the end goal is a worthy one.
Imagine an area of the size of surface of the earth. Put 30k objects of the size of motorcycle on it. There are currently something like 200 million motorcycles in the world. Seems like plenty of space, doesn't it?
But wait, you don't actually need to restrict yourself to one spherical surface. You can place your 30k sattelites anywhere in a Earth's-surface-sized spherical shell that's 400 miles thick. You could literally put those 30k sattelites in a single straight line above some point on the earth, and still have 350 miles of shell thickness left.
Congestion on low Earth orbits is really a non issue, especially if the satellites are designed to deorbit after their useful lifetime.
Surely you have something to back this up other than a super weak analogy?
Lets pick your assumption apart - there are 200 million motorcycles in the world, but there are only an esitmated 35 million km of roads. That means there are approximately 175 meters of road per motorcycle. That seems pretty reasonable, even if they were all on the road at the same time. But they aren't alone - there are also approximately a billion vehicles. That takes it down to about 35 meters per vehicle. Thats without factoring in the loss in space for intersections, the drop in capacity at each of those intersections, or addressing the fact that most of those vehicles are clustered in high density urban areas.
Now we are lucky that not all of those vehicles are on all of those roads at the same time, because they can be removed and parked.
It is certainly possible to overlay orbits at different altitudes, but then you have to manage spectrum congestion, because those 30k satellites in a straight line above some point on the earth are useless unless they can actually communicate with ground stations.
Not all orbital slots are created equal either - some have better coverage of human habitable areas than others.
Oh, and there is the small matter that each time one of those 30k satellite degrades, it becomes between two and hundreds of thousands of devastating fast moving particles that can destroy other satellites (don't believe me? look up the number of windows that have been destroyed by paint flecks in orbit).
Don't take my word for it, read up on NASA's articles and quarterly publications on orbital congestion.
Difficulty: the motorcycles are moving at 20,000 miles an hour. And if two of them ever collide, not only will the motorcycles be totaled, but so will the road.
I should have been more clear. By "congestion", I was not just considering the number of satellites, but the number of artificial objects in total regardless of size, which by many estimates is in the millions.
Starlink satellites are mostly put into deliberately low orbits to guarantee they degrade over the course of a few years and burn up in the atmosphere. This is not at all the same as putting the same number of normal LEO satellites in orbit from a space junk perspective.
If you remember the issue where SpaceX refused to move for the ESA sattelite. I think this a situation where space and air traffic control have to become more automated. This is like the beginning of an iteration in construction like the ones that happen in suspension bridges. The hope is that the Tacoma narrows scenario doesn't end up too costly. One of the fun things is that with this many sattelites you have a pretty good distributed orbital debris tracker.
That's absolutely true. However, even then, space debris is already a major concern[0], and let's not forget the recent debacle with the ESA[1]. What happens when you pump 30,000 more satellites up there?
GNU/Linux is already available on handhelds that can connect to traditional networks but this will make things trivial. All android will be is a runtime that will slowly die as google attempts to weaponise it.
This is complete nonsense. There is not enough demand for all those satellites and capacity, so it's very hard to understand why they are making such a filing.
As of today, there is probably ~ 1 Tbps of usable capacity in space (wordlwide). With such a big constellation they could provide 100x this capacity, but who is going to buy it? How are they going to compete with fiber-delivered Internet in big cities? They cannot offer the same QoS, so you do not want space delivered Internet if you can have it through terrestrial networks.
First off, these constellations will (in theory) have fewer tradeoffs vs terrestrial networks.
At only 350-800 miles up, they are much lower latency than current satellite based Internet. These in theory add similar levels of latency as DOCSIS - and tolerable levels to most internet based applications.
It won't replace fiber for very latency sensitive applications - but being able to deliver a tremendous amount of bandwidth virtually anywhere with a penalty of only ~7-10ms is very compelling.
Secondly, in rural areas, everyone has a view of the sky. Very few have compelling fiber, or even cable/DSL options.
This could dramatically reduce the cost of backhaul services to ground based cellular stations - and improve connectivity in a variety of other ways.
I don't think that latency is a decisive factor for most people / networks (as long as its below certain threshold). Most people will tell you it's a price-based decision.
On the other hand, roughly 85% of the population worldwide is already covered by 3G or 4G networks. I am not sure that space will be able to offer lower prices for backhaul if the infrastructure is already there. A fiber PoP / tower has much lower OPEX than launching, operating a constellation of satellites.
>On the other hand, roughly 85% of the population worldwide is already covered by 3G or 4G networks.
First, and less importantly, source. Because phone companies love to toss out coverage numbers that are pure, unadulterated bullshit. But second that's less important because 3G and 4G are both completely uncompetitive for broadband. LTE is basically the minimum there. And awful data caps and congestion remain a serious issue all over the place.
>A fiber PoP / tower has much lower OPEX than launching, operating a constellation of satellites.
Is it? I mean, seriously. A lot of SpaceX's plans are based on estimated launch costs via Super Heavy, so necessarily somewhat speculative. But the target is sub-$100/kg, and at least in terms of fundamentals (like cost of fuel, Raptor effectiveness, estimated cadence) that doesn't look ridiculous. Each Starlink satellite apparently has a mass over around 230 kg. Add a bit of extra mass for some bigger sats in the mix, double the launch cost to be a bit more conservative, and we're still talking a launch cost per satellite of well under $60k a piece. They necessarily have everything needed to operate their design lifetime, because there is no maintenance that can be done on them. They're operating low enough that solar weather shouldn't be too constant a concern, and of course will never face OPEX costs like vegetation clearing or weather damage or corrosion or the like. Also doesn't face NIMBY tower politics either. Is it actually more expensive then running fiber and building/maintaining a tower? For very high density locations probably, but there's a pretty sizable population in the US alone for whom it seems on napkin math it could be very competitive.
For the source, see [1]. "The coverage gap– those living outside of areas covered by mobile broadband networks4–continues to decrease below the 1 billion threshold and now stands at 750 million people. Since 2014 the gap has more than halved, from 24% to 10% of the global population." It's actually 90% covered by 3G and 4G.
Regarding the costs, I am not sure where your 60k/sat is coming form. Currently, launch costs for SpaceX are in the millions (60 commercially, let's say 30 millions cost), and they are launching 60 satellites at a time, so at least 500k/sat.
I am not sure that they will be able to cut the launch costs by a factor of 10, even considering resusability, larger launch capacity, etc, etc.
It is completely clear where the 60k number is coming from. Whether you agree with it or not (he did qualify it as speculative) if you don't know where it came from then you must have replied to his post without actually reading it.
You are right, maybe I should have said where the sub-100$/kg comes from.
In the first Starlink launch, with 60 satellites of ~230 kg each, the cost of launch was ~$30M (considering that the "retail price" of a Falcon 9 launch is $60M, I guess that half the cost if a good assumption). Thus, the cost per kg was $30M / (60 * 230 kg) ~ $2,000/kg. I just don't see how they are going to achieve a x10 reduction in cost, even considering reusability, larger masses, etc, etc.
Yeah, I guess we'll see. Economies of scale and complete many-times reusability of both stages sound like plausible multipliers to me, but it's all speculation until the thing flies a few times and they get a mature production line up and running.
For rural coverage, fiber builds are _very_ expensive.
For cities, you're right - 5G + high frequency wavelengths is probably cheaper - but not by that much assuming SpaceX's numbers are accurate.
The problem is 85% of the population is absolutely not covered by high frequency spectrum - tower/DAS coverage is optimized for sub 1Ghz bands. We would need billions upon billions of additional tower builds to deliver an adequate experience on spectrum that basically requires line of sight. Even in cities. Some of that has been invested already (e.g. VZ can reuse their Fios plant in many locations), but certainly not everywhere.
roughly 85% of the population worldwide is already covered by 3G or 4G networks.
Consider the reverse.
As internet access becomes a staple service for modern living, one’s options for location are restricted to where internet is available: 3G coverage at a barely functional minimum.
Want to solve the housing crisis? poverty? Bring at least LTE performance (100Mbs better) everywhere at a competitive flat rate. Today, many options are untenable for want of a decent data rate.
How do current ground wired customers like their internet provider? On the whole, they don’t.
What percentage of the world population has internet access of any kind? of land area? of anything over 1Mbps? Huge untapped markets there.
I see this building up like onset of digital photography: nobody expected supplanting film & paper, yet the whole world switched seemingly overnight. Kodak went from most valuable brand to “who?” fast.
Once people see Starlink working fast, reliably, and anywhere they’ll switch en masse. Personally, as a 100% telecommuter, the only thing stopping me from moving where I want is internet connectivity.
One billion customers divided by 40,000 Starlink satellites is a very manageable 25,000 users each. Sometime please run the numbers on supporting 100Mbps for each, plus latency. Given that, I expect a huge migration of customers.
Yep. This is the kind of "create new markets" move that seems to not make sense, but will be a spectacular success if it pans out. And no one else will be able to compete. There is only one company in the world that will be capable to economically launch 30,000 satellites in a single-digit number of years.
The higher altitude satellites are better for serving cities because the total bitrate gets averaged over more area - so you can have a hotspot of users in the city as long as there aren't many users in the surrounding ~600km circle.
Cities are spaced ~1200km apart in much of the world.
Well, you are saying that you have a fixed bitrate per beam to divide among all the users in the area. The bigger the area the more users. And once you reach saturation, that's it. Also, if there is a very populate city inside the beam footprint, you are only going to be able to serve to a very limited amount of people in there.
Keep in mind current cell phone coverage is spotty, with many unserved or underserved areas, and towns where Verizon is clearly better than AT&T, or vice versa.
Running that kind of cell tower network might be cheaper, but running a cell phone network that spans every inch of the globe would be an engineering and logistics feat that would almost be impossible to profit off of.
Even maintenance-wise, servicing a cell tower requires keeping a bunch of contractors around to drive to drive, fly, or even boat to remote areas, and climb a hundred feet once they get there. With starlink, they can just piggyback replacement satellites on customer payloads.
Also, there are around 4 million cell towers in the world to achieve spotty coverage, whereas starlink is targeting 40k satellites for global coverage. I'd have to believe that launching 40,000 similar payloads via similar methods has huge vertical integration advantages over the extremely heterogeneous approaches we use to put up those millions of towers.
Sorry, I think I was misunderstood. Cell towers that work over starlink (i.e., use starlink as an uplink method) may be cheaper than cell towers that use a landline for an uplink.
So the ideal situation would be Starlink for remote areas, and in medium to higher density areas put up towers that talk to Starlink, then can speak to local receivers using terrestrial radio technologies.
>How are they going to compete with fiber-delivered Internet in big cities?
Who is this mystery company building fiber-delivered internet in big cities? You must not be talking about the US, because:
#Google
#CenturyLink
#Verizon
Were the major players building out, and all have stopped. That's ignoring the tens of millions of people who don't live in the city and have 3Mbps DSL as their best case scenario. And that's ignoring the mess in Australia and Canada. There will be PLENTY of demand for high speed, low latency satellite.
I am not talking fiber to the home, I am talking a fiber backed infrastructure, like cell towers and such.
If the solution was to provide Internet wirelessly, we would have that right now. The cell phone infrastructure provides much higher data-densities than any satellite network will ever be able to provide.
I'm not following your point. The whole reason Starlink is being launched is for home broadband. Every cell provider in existence has ridiculously low caps - starlink is supposed to be uncapped. You're the guy telling someone he doesn't need a dump truck for his landscape business because we have public buses.
You could provide home broadband using a wireless, mobile-network like system if you wanted to.
It is not happening because it's very difficult to do that with the spectrum allocated for mobile services. A network of towers can provide higher capacity densities than what SpaceX's 30k satellite systems will be able to.
In big cities, they will fall very sort of the capacity required. The current 4,409 satellite proposal has beams of ~700 Mbps, and each of those covers a surface larger than the whole NYC metropolitan area. The capacity scales linearly with the number of satellites in line-of-sight (LoS). Right now, with the 4,409 satellite system you can have ~40 satellites in LoS, so they can provide ~28 Gbps of throughput to the whole city. Even with 30k satellites, they will only be able to provide ~196 Gbps. That is not enough to offer any uncapped data plan.
> You could provide home broadband using a wireless, mobile-network like system if you wanted to.
Why do you continue referencing one of the most densely populated cities in the world as the unit of measure? This is for un-served and under-served areas which is half the globe. Of course they aren't building this to displace NYC's current ISPs. They clearly ran financials before investing... you seem like you're just trying to find a reason to say it's stupid without bothering to consider that someone smarter than you MIGHT have done a little research before spending a couple billion dollars.
My PhD research is on how to use space and aerial networks to serve the unconnected and undeserved. I have been studying these networks (and things like Google's loon, mmwave, etc) for the last 4 years. I have simulated these networks in detail, assessed their capabilities, and ran a lot of financials myself.
According to my analyses, there needs to be a disruptive change in how we receive broadband Internet to close the business case (that's why I keep talking about Displacing ISPs). The unconnected and undeserved broadband market plays great in the US and Canada, but in the rest of the world, there is simply not such a market. At best, you are going to try to provide backhauk for a cell phone network or some other kind of access point. The whole 3B of unconnected people rethoric is very slim. Those people can afford to pay very little every month for connectivity, so it's very hard to make money out of it (even to offset your marginal costs).
In terms of performance, there are diminishing returns on each extra satellite that you launch. The ones already up there are already satisfying a lot of the demand, so each newly launched satellite is active for a shorter and shorter percentage of time.
Starlink is a (in my opinion) very oversized system. I do not think they will launch the full 4,409 constellation (as it stands today), and I am pretty confident they will not launch a 30,000 satellite constellation.
Finally, LEO networks are not the most cost effective space networks. GEO HTS and MEO networks have lower CAPEX/GB.
~45% of the world's population lives in rural areas. That's a lot of potential customers, myself included; I live 30 minutes north of a US state capitol, and my only option for terrestrial internet at home is 1.5mbps DSL (which I currently pay $70/mo for), or my cell carrier (which is spotty and unreliable in my house), so I'd definitely consider SpaceX Internet.
Even developed areas in the US typically are stuck with one broadband provider. If you do have a second one, everywhere I've lived it's been DSL service and I've never heard anything good about it aside from being cheaper than Comcast.
Exactly, that is the real problem, the lack of competition from terrestrial networks. In the US the ISPs operate as a monopoly. In most big cities in Europe you have 3/4 providers to choose from, which reduces costs and increases the quality of service.
They deliver better latency than current sat-based internet. They also want to use fixed phase array antennas instead of large dishes or tracking dishes. This is a new kind of service that has not existed before.
DSL modems are a similar technology leap - they used to contain high end FPGA's, DSP chips, and high speed analogue to digital converters, and they'd cost hundreds of dollars.
Now a cheap DSL modem is $10, and it's all a single SoC.
Phased array satellite modems will follow the same price curve.
I hope that ends up happening, it will be a bigger innovation than all these constellations!!
I don't think it will happen in the next 3 years though. I think that most of the manufacturers are focusing on the high-end market, and first terminals will be in the thousands of dolars.
I think it will happen, because phased array systems are being developed for 5G, next-gen WiFi, etc. Lots of players in the ASIC design game are designing similar hardware, and in the volumes spacex presumably wants to order, I think they'd be totally happy to spin a chip putting together a few hundred phase shifters and amplifiers in-front of a regular DSL modem SoC.
Another point which may be missing from the analysis: if I were Musk and I wanted 3000 satellites in orbit I'd probably also make a filing for 10x that and see how much I can get / how much I will need to meet demand.
https://www.reddit.com/r/SpaceXLounge/comments/dge0xk/we_nee...