SpaceX Says It Will Deploy Satellite Broadband Across US Faster Than Expected (arstechnica.com) 60
SpaceX says it plans to change its satellite launch strategy in a way that will speed up deployment of its Starlink broadband service and has set a new goal of providing broadband in the Southern United States late next year. Ars Technica reports: In a filing on August 30, SpaceX asked the Federal Communications Commission for permission to "adjust the orbital spacing of its satellites." With this change, each SpaceX launch would deploy satellites in "three different orbital planes" instead of just one, "accelerating the process of deploying satellites covering a wider service area." "This adjustment will accelerate coverage to southern states and U.S. territories, potentially expediting coverage to the southern continental United States by the end of the next hurricane season and reaching other U.S. territories by the following hurricane season," SpaceX told the FCC. The Atlantic and Pacific hurricane seasons each begin in the spring and run to November 30 each year.
SpaceX said it already planned to "provide continual coverage over northern states after as few as six more launches," but said it needs a license modification to speed up deployment in the Southern U.S. SpaceX's filing stresses the importance of quickly getting service to parts of the U.S. where broadband coverage is limited. "With this straightforward adjustment, SpaceX can broaden its geographic coverage in the early stages of the constellation's deployment and enable service initiation to serve customers earlier in the middle latitudes and southern-most states, and critically, those often underserved Americans in Hawaii, Puerto Rico, and the U.S. Virgin Islands," the company told the FCC.
SpaceX said it already planned to "provide continual coverage over northern states after as few as six more launches," but said it needs a license modification to speed up deployment in the Southern U.S. SpaceX's filing stresses the importance of quickly getting service to parts of the U.S. where broadband coverage is limited. "With this straightforward adjustment, SpaceX can broaden its geographic coverage in the early stages of the constellation's deployment and enable service initiation to serve customers earlier in the middle latitudes and southern-most states, and critically, those often underserved Americans in Hawaii, Puerto Rico, and the U.S. Virgin Islands," the company told the FCC.
I have to wonder (Score:3, Insightful)
if this will cause some serious price reductions in ISP prices, especially in areas with only 1 choice of providers.
Re:I have to wonder (Score:5, Informative)
Probably not soon (Score:5, Insightful)
Musk has said [twitter.com] that each launch of 60 Starlink satellites adds about 1 Tbps of usable capacity, meaning roughly 16 Gbps per satellite. This has to be shared among all users in line of sight, so if you want 100 Mbps then you'd better hope there are no more than 159 other like-minded users in the satellite's footprint (depending on their oversubscription policy). The number of satellites in each orbital plane will determine the speeds available to subscribers, with denser areas seeing more congestion.
Initially, SpaceX is planning [spacenews.com] 72 orbital planes with 22 satellites per plane. I'd expect to start with you'll see maybe 3-4 satellites visible at a given time, so if they offer 100 Mbps plans and oversubscribe 10 to 1, then that's maybe 10,000 subscribers per city at best before peak speeds start to suffer significantly. That's not going to threaten Comcast or any big-city ISPs, but could be fantastic for rural areas with limited choice. However, as the initial generation will not support satellite-to-satellite links, you'll still have to hope there's a ground station somewhere in your general vicinity.
Of course the final constellation will be much larger, maybe even 7x those numbers. But I'm having trouble seeing how Starlink will provide significant competition to ISPs in any moderate sized city.
Re: (Score:3)
Oversubscribing 10:1 would be way more than normal. I know that at the biggest data party here in Norway they were averaging about 20 Gbps for about 5000 people so 4 mbit/s average. And that's all people actively using their computer, though I guess many are gaming and doing other things that aren't that bandwidth intensive. But like, I'm reading slashdot right now and not really using any practical amount of bandwidth. Assuming they have some kind of peak/off-peak caps or pricing to shift some of the load
Re: (Score:2)
Re: (Score:3)
But I'm having trouble seeing how Starlink will provide significant competition to ISPs in any moderate sized city.
Quite simple. Apart the usual 'limited number of providers', they have another advantage: once those satellites are up, the cost for infrastructure for any subscriber is basically close to zero.
They don't have to install or maintain any land wires, or lease them from a 3rd party (common in Europe, not sure about the USA). The cost for launches have been made anyways, the cost per subscriber is some modem (< $100) and some administrative fees. Bandwidth itself is basically as good as free.
Also, your overs
Re: (Score:2)
The cost for launches have been made anyways, the cost per subscriber is some modem (The reported target price for the base station with antenna was $200, and they were having trouble actually meeting that target. Phased array antennas aren't cheap to fabricate, apparently.
Re: (Score:2)
Argh. Missed the '<' in the quote. Let's try that again.
The cost for launches have been made anyways, the cost per subscriber is some modem (< $100) and some administrative fees.
The reported target price for the base station with antenna was $200, and they were having trouble actually meeting that target. Phased array antennas aren't cheap to fabricate, apparently.
Re: (Score:2)
Doesn't really matter to Comcast how cheap it is for Starlink per subscriber, or how cheap their retail rates are, if they can only steal 10,000 customers max from the New York market. That wouldn't hurt them, nothing worse than a little bad press over the price difference perhaps, and I'm pretty sure Comcast is immune to bad press by now.
Agree that 10 to 1 is pretty generous, yeah. In Australia the government-owned network operator charges retail ISPs about $38 per (100 Mbps) customer per month, just for t
Re: (Score:3)
My guess is SpaceX will be marketing harder towards commercial users than residential users.
Ships, planes, trains, financial trading, transport trucks, and remote businesses/offices will benefit immensely from a cheaper satellite-based internet provider than any current satellite or cellular-based provider. Commercial per-device data use will probably be lower than residential use, and commercial customers will probably pay more too.
This is just my guess though. After seeing these numbers, I don't see how S
Re: (Score:2)
They won't be targeting remote ships & planes initially. As the first wave of deployed satellites only support communication to & from ground stations, then they'll only be able to operate when there's a ground station within 1000 km or so, which limits service over oceans a lot. But you're right that those type of customers will be extremely interested - their only options are pretty slow & expensive right now.
Re: (Score:2)
Actually the round-trip latency to a Starlink satellite in e.g. 500 km LEO is under 4 ms - vastly better than the geosynchronous satellite service you're probably thinking of, which is more like 250 ms. And they'll have plenty of bandwidth - so long as they limit their subscribers in any given footprint.
Re: (Score:2)
What exactly does anyone wants 100 Mbps for, though? Personally I wouldn't pay an extra dollar for anything over 10 Mbps, because there's nothing of use I can do with it.
Re: (Score:1)
Re: (Score:2)
if this will cause some serious price reductions in ISP prices, especially in areas with only 1 choice of providers.
Don't count your chickens before they're hatched. Comcast, AT&T, Verizon's CEOs and the other owners of US telecomuncations fiefdoms have long since made a trip to the White House to make the case that: Mr President we are getting **competition** ... something must be done!!!
Re: (Score:2)
This is probably actually great for them. This will mean AT&T can pretty much walk away from ever having to provide broadband to its rural customer base or they can contract with SpX to deliver service to those users.
It'll help with rural 5g deployment since you can slap a starlink antenna on top of a cell tower and provide broadband links without having to worry about upgrading the backhaul to the tower.
Granted I'm in the unusual position of having 3 home gigabit provider choices, but i think people a
Re: (Score:2)
Granted I'm in the unusual position of having 3 home gigabit provider choices
Let me guess: Google Fiber, Spectrum & AT&T?
Re: (Score:2)
Muni Fiber, Comcast cable, and Centurytel Fiber.
At one point comcast had 2gbps fiber as well, but htat doesn't seem to be an option right now
Re: (Score:1)
That city provided protection from another IPS cant go up into orbit
Here's hoping (Score:2)
if this will cause some serious price reductions in ISP prices, especially in areas with only 1 choice of providers.
In some areas it might. Probably not so much in dense urban areas with solid wired connections and copious cell towers. Homes that live in dead zones or which aren't reached by high speed internet are a real thing and hardly uncommon.
I'll certainly take a look at it should it come to full operation. Right now my only realistic options for internet are Comcast or LTE wireless. Both are fairly expensive and problematic as you might guess. I'm not likely to get a second competitive wired option where I li
Re: (Score:2)
Re:I have to wonder (Score:4, Informative)
if this will cause some serious price reductions in ISP prices, especially in areas with only 1 choice of providers.
The ironic thing is: the real point of Starlink is to provide $100 million/year ISP plans to a handful of customers who already have the best choice of providers.
For those wondering how Starlink will "fully fund SpaceX", they can knock a few milliseconds of the best current latency between e.g. New York and London, enabling faster reaction to market changes and news. If you're unfamiliar with the amount of money involved here, a private undersea cable was laid between NY and London just to get a shorter cable route than the best previous, reducing the latency from 65 ms to 60 ms, for $300 million, and that was massively profitable. Starlink will offer a much better then 5ms gain.
Re: (Score:2)
Yeah i suspect there's a lot of revenue possibility here. London and NY might be the big two at the moment, but i'm sure the reduction in trip time between the markets in shanghai and tokyo will be very valuable too
Re: (Score:2)
The stock exchange theory has been speculated about by third party writers, but to my knowledge SpaceX has yet to indicate if they're planning to pursue that strategy. On the surface it may appear an obvious strategy, but there may be non-obvious difficulties.
Re: (Score:1)
How much is this supposed to cost? (Score:2)
If it's not too high, I'll give it a try. Anything to give the current illegal Comcast monopoly in my neighborhood the finger.
I dont care about latency that much, I'm not a gamer.
Re:How much is this supposed to cost? (Score:4, Interesting)
Probably around a bill a month. That's what existing satellite services cost, but with a second of latency and abusive throttling of streaming connections at all hours (i.e. whether they are necessary at the time, or not). Plus zero of them have self-aiming equipment for RVs. If Starlink solves all three of these problems, they could actually charge MORE than Viacom and still get plenty of subscribers.
Re: (Score:2)
I think this is probably their best strategy. Go after rural customers first because it won't be hard to beat what's currently out there. They are going to have to have a pretty good service to beat companies that have cable.
There's a lot of bandwidth in a cable TV cable. The last couple years the Canadian cable providers have really started to be competitive because of so many third party providers that they cable owners are required to lease their lines to. We used to have low speeds and small data caps,
Re: (Score:3)
Re: How much is this supposed to cost? (Score:3)
That's about right.
Because light in a vacuum is remarkably faster than light in a fibre (much less electrons on a wire), total latency from distant points on the globe will actually be lower with Starlink than terrestrial.
Re: How much is this supposed to cost? (Score:3)
20 metres ? That IS a low orbit. Why do they need such big rockets ?
Re: (Score:3)
20 metres ? That IS a low orbit. Why do they need such big rockets ?
There's a lot of drag at that altitude.
Re: (Score:2)
How? (Score:2)
Not appart, actually (Score:2)
Can they steer those satellites to that degree?
For that: yes, they can steer satellites. They are actually reaching final orbit by their own propulsion means.
Though probably there are going to be multiple launches and each launch is going to mostly launch those sattelite whose orbit benefint the most from the launcher's initial trajectory. So do not expect the individual sattelite each making giant 60 degree of orbit adjustement. (Too much propellant- and energy- costly. Better launch them all in the general vicinity and only do small adjustements).
I assume 3 planes are 120 degrees apart
Not
Re:How? (Score:5, Informative)
You're assuming that one launch would cover the entire planet, which is incorrect.
Previously they were approved to put satellites in 24 orbital planes which would have been separated by 15 degrees. They want to increase that to 72 orbital planes which would only be separated by 5 degrees.
The key is 'orbital precession'. (Score:3)
Because Earth is not a sphere, items not in equatorial or polar orbits 'precess', or the RAAN ('longitude' of the orbit) drifts westward. Objects in lower orbits precess faster.
So, spaceX will launch into the most eastern of the planes, and quickly lift satellites meant for that plane up into the working orbit. Then they'll wait until the sats in the low parking orbit precess into the next plane west, and lift those due for that plane up. Then wait again until the rest of them are in the third plane, and th
Re: (Score:2)
3 orbital planes one launch? Can they steer those satellites to that degree? I assume 3 planes are 120 degrees apart
The satellites do not have *that* much delta-V, but the second stage might.
The orbital planes are inclined from each other at a much smaller angle.
Can't come too soon (Score:2)
Hurricane season? (Score:2)
Why do they mention hurricane seasons?
Re: (Score:3)
They'll get some free advertising if they're the only working network link after a natural disaster. And it will give them a weapon to try and fend off the old guard ISP controlled political apparatus on the ground. (Why are you denying this essential service used by rescue workers a faster roll out? People are dying.)
Re: (Score:2)
Because rocket launches can't happen in hurricanes...?
Re: (Score:3)
Probably because local infrastructure can be offline for extended periods during and after an extreme weather event. But if you have starlink you can with minimal effort get communications back online (minimal meaning a small power source like a portable generator, a starlink modem, and a wireless router and you now have dozens or more than can use their cellphones).
Re: (Score:2)
let me reinterpret that for you (Score:2)
Unfortunately there's two ways to parse that. I shall choose to optimistically read that as a surplus of supply rather than a deficit of demand.
Re: let me reinterpret that for you (Score:2)
SpaceX has always been clear that Mars is their goal and that terran revenue streams are to fund that goal.
Both Starlink and e2e are major upcoming revenue streams. Block 5 was developed, in part, to get 14,000 birds in the sky. F9 can only do 60 at a time but Starship Cargo should be able to drop off 420 at a time. 48-hour turnaround on Block 5 should help but full reusability on Starship should be much faster (yay, methane). Anyway, six Block 5's isn't /too/ much work for SpaceX. They don't want idl
Re: (Score:2)
Re: (Score:2)
In other news - "SpaceX announces fewer customers than launches."
Unfortunately there's two ways to parse that. I shall choose to optimistically read that as a surplus of supply rather than a deficit of demand.
If I were them, I'd intentionally accept fewer customers for a while, thus avoiding advertising costs and giving time to build a customer service organization, something SpaceX effectively doesn't have. They're going to have to deal with phone calls for billing problems, antenna siting problems, and the usual steady stream of complainers.
Sure you can hire a company to do that for you, and they may, but even that takes ramp up time and there may actually be legitimate problems in amongst all the user errors
damn exciting! (Score:1)
Re: damn exciting! (Score:2)
All the border areas will be well-covered by the next six launches. Northern US and Southern Canada. You'll have to wait if you're up by the Arctic Circle.
Re: (Score:2)
It took Agriculture and Agri-Food Canada to get involved for rural farmers north of the border to get connected. And pay almost double the cost of a reasonable urban plan for the privilege of a measly 1.5Mbps.
How open do you think Elon would be to including Canada when he covers Alaska? The market up here is ripe for a shakeup.
I'm sure SpaceX would love to do that but based on Canada restricting access to US based satellite providers, I am dubious that it will be permitted.
Faster Than Expected (Score:2)
... so they will be in a higher orbit ?
Re: (Score:2)
Depends on how you interpret it. A lower orbit is faster than a higher one. But a higher orbit requires a higher delta v to get there.
Re: (Score:2)