Stories
Slash Boxes
Comments

SoylentNews is people

posted by Fnord666 on Wednesday March 25 2020, @05:01AM   Printer-friendly
from the space-is-becoming-a-crowded-place dept.

SpaceX gets FCC license for 1 million satellite-broadband user terminals

SpaceX has received government approval to deploy up to 1 million user terminals in the United States for its Starlink satellite-broadband constellation.

SpaceX asked the Federal Communications Commission for the license in February 2019, and the FCC announced its approval in a public notice last week. The FCC approval is for "a blanket license for the operation of up to 1,000,000 fixed earth stations that will communicate with [SpaceX's] non-geostationary orbit satellite system." The license is good for 15 years.

[...] One million terminals would only cover a fraction of US homes, but SpaceX isn't necessarily looking to sign up huge portions of the US population. Musk said at the conference that Starlink will likely serve the "3 or 4 percent hardest-to-reach customers for telcos" and "people who simply have no connectivity right now, or the connectivity is really bad." Starlink won't have lots of customers in big cities like LA "because the bandwidth per cell is simply not high enough," he said.

SpaceX's main Starlink constellation competitor is running out of money

OneWeb, the only pressing competitor facing SpaceX's Starlink satellite internet constellation, has reportedly begun to consider filing for bankruptcy shortly before the London-based company completed its third dedicated launch.

Following the completion of its first full 34-satellite launch with a Russian Soyuz rocket on February 7th, OneWeb managed to complete a second launch on March 22nd just a few days after Bloomberg revealed its bankruptcy concerns. OneWeb now has 74 ~150-kg (330 lb) satellites in orbit – roughly 11% of its initial 650-satellite constellation. Like SpaceX, OneWeb's goal is to manufacture and launch an unprecedented number of high-performance small satellites for a per-spacecraft cost that would have previously been inconceivable.

[...] Requiring numerous revolutions in satellite manufacturing, antenna production, and launch vehicle affordability, as well as a vast and complex network of ground terminals, numerous companies have tried and failed to rise to the challenge over the decades. Original Globalstar, Teledesic, and Iridium constellations all raised more than $10 billion in the 1990s under the promise of blanketing the Earth with internet from space. All wound up bankrupt at one point or another.

See also: The true impact of SpaceX's Starlink constellation on astronomy is coming into focus

Previously:
SpaceX Seeks Approval for 1 Million Starlink Ground Stations, Faces Pentagon Audit
SpaceX and OneWeb Clash Over Proposed Satellite Constellation Orbits
OneWeb Joins the Satellite Internet Gold Rush this Week
OneWeb Launches its First Large Batch of Broadband Satellites, Plans March Launch and April Break
How Does Starlink Work Anyway?


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 3, Insightful) by Immerman on Wednesday March 25 2020, @01:55PM (15 children)

    by Immerman (3985) on Wednesday March 25 2020, @01:55PM (#975456)

    Sounds like per-satellite max throughput bandwidth is 20Gbps - assuming you could use all of that for customers on the ground (I would assume the reality would be most of that goes to inter-satellite relays) That would mean a single satellite coverage area "cell" could only support at most 20Gbps/100Mbps = 200 simultaneous people at 100Mbps. And likely only a fraction of that.

    Potentially a few neighboring satellites could overlap their service area to increase that a bit, provided the surrounding area was empty. But with 12,000 satellites covering most the Earth, the average cell size is going to be in the ballpark of 200e6mi2/12,000 = 17,000square miles (roughly 130 miles across).

    So, *really* not a viable mass alternative in urban, or even suburban areas. Perhaps an option for a relative handful of people that want high availability or low latency, and presumably are willing to pay for it. I imagine hospitals operating remote surgery centers might benefit, and of course scum like high-frequency traders are probably a primary funding source. And those who are willing to pay a premium for not dealing with their local monopoly. Mostly though I suspect it will go to low-bandwidth connections (1Mbps) for people who don't really have any other options.

    Starting Score:    1  point
    Moderation   +1  
       Insightful=1, Total=1
    Extra 'Insightful' Modifier   0  
    Karma-Bonus Modifier   +1  

    Total Score:   3  
  • (Score: 2) by takyon on Wednesday March 25 2020, @02:31PM (1 child)

    by takyon (881) <takyonNO@SPAMsoylentnews.org> on Wednesday March 25 2020, @02:31PM (#975467) Journal

    I don't think they can stop people from trying to use it in suburbs. It's usable since the sky can be clearly seen and the instructions are "plug in and point at sky, in either order". 5 homes in a row might get it for an RV but try to use it while at home. Unlikely, but it could happen.

    Maybe it just won't be very popular with people who already have a half decent connection, and that will limit the problem.

    --
    [SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
    • (Score: 2) by Immerman on Thursday March 26 2020, @02:29AM

      by Immerman (3985) on Thursday March 26 2020, @02:29AM (#975695)

      Actually, it'd be pretty easy to stop: just don't send or acknowledge any data associated with rural accounts trying to connect within an urban area. Every satellite knows what area it's currently serving, and your ISP always knows whose account to bill for every byte of data.

      Even if they didn't though, it'd be fairly pointless to do. That basic service package that gets you 100Mbps in the far end of nowhere? You get within 70 miles of a city and you'd have to start sharing that bandwidth with many thousands of other people instead of only hundreds. Your performance is going to tank. Especially since you'll almost certainly be sharing whatever bandwidth is left over after the high-dollar premium accounts have tranceived their fill.

  • (Score: 3, Interesting) by deimtee on Wednesday March 25 2020, @03:03PM (12 children)

    by deimtee (3272) on Wednesday March 25 2020, @03:03PM (#975480) Journal

    It's not quite as bad as that. The maximum orbital inclination is supposedly 70° so the coverage will be higher near the equator, with uncovered areas near the poles. Also, not many connections max out full time - it might work out to 200 customers per cell by your math, but every existing telco would happily sell 2000 connections per cell and cheerfully claim they are all 100/100Mbps.

    Aside from that though, while 100/100 is a nice connection, I know people in remote areas who would be very happy with low-latency 5/1Mbps if they could get it. Current satellite internet is ok for downloads once it gets going, but it's really poor for surfing with the lag being applied to the hundreds of consecutive requests bloated websites make.

    --
    No problem is insoluble, but at Ksp = 2.943×10−25 Mercury Sulphide comes close.
    • (Score: 2) by Immerman on Thursday March 26 2020, @02:00AM (11 children)

      by Immerman (3985) on Thursday March 26 2020, @02:00AM (#975675)

      I was simply showing that even 100Mbps in the suburbs is not realistic. Even in most semi-rural areas it would be a stretch. And sure, an ISP would sell 10x that capacity, mostly without trouble - but if there's 200 people that want to actually use the speed at the same time you start to run into problems. 200 people in 17,000 square miles? Not that hard. Even 100 square miles would hit that limit almost instantly if it intersected a metropolitan area.

      Hey, I'm not one to dis low bandwidth - I only recently upgraded from 1Mbps to 20 and, aside from downloading large files, the functional difference is not that great. Being able to offer 20,000 people in a poorly-served area simultaneous 1Mbps is a wonderful thing, especially since you can probably actually deliver those speed on demand to 100,000 or more. It's just not something valuable to most people, because most people live near an urban area that would be grossly underserved if more than a tiny fraction of people used the service.

      • (Score: 2) by takyon on Thursday March 26 2020, @03:43AM

        by takyon (881) <takyonNO@SPAMsoylentnews.org> on Thursday March 26 2020, @03:43AM (#975715) Journal

        I don't see SpaceX targeting very low bandwidths, even if it is an improvement for some. Maybe 50 Mbps at the lowest. Hopefully we will be able to stop speculating about this soon, but the service will probably be delayed by coronavirus like everything else.

        It would be helpful if new versions of the sats could do better than 20 Gbps total. The current versions don't have sat-to-sat links either. But the more pressing design issue right now might be a sunshade [spacenews.com] to lower the impact on astronomy.

        --
        [SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
      • (Score: 2) by deimtee on Thursday March 26 2020, @05:16AM (9 children)

        by deimtee (3272) on Thursday March 26 2020, @05:16AM (#975748) Journal

        Actually, another way to look at it would be to count how many satellites are over the country at any one time and assume you could support n customers where n = (no. sats) x (bandwidth / (bandwidth per cust)) x (capacity factor)

        no. sats = 12,000 / (2 x 10) : half in N hemisphere, USA is about 1/10 of circumference of the world.
        b/(bpc) = 20
        CF probably 10

        = 120,000 customers in USA.
        @ $50 p/month = $150M per year return from USA.

        ----

        19/20ths of the time the satellites won't be over the USA. Somewhere up to half the time they will be over another country. I think the profitability of Starlink is really going to depend on being able to sell in other countries. (I am actually more interested in AU, cos I live here. About the same size, and could probably supply just as many customers.)

        --
        No problem is insoluble, but at Ksp = 2.943×10−25 Mercury Sulphide comes close.
        • (Score: 2) by Immerman on Thursday March 26 2020, @02:09PM (8 children)

          by Immerman (3985) on Thursday March 26 2020, @02:09PM (#975869)

          Don't forget the high frequency traders - they made it worth laying new trans-atlantic cable to shave a few milliseconds off lag times. They'll probably be major profit centers for Starlink as well.

          I think you made a mistake in your math as well, at least in the intermediate step. Given, say 20Mbps bpc, - b/bpc is = 1,000 customers per satellite. Even 100Mbps = 200 cust/sat
          x (CF of 10) x (600 satellites) = 1.2 to 6 million customers
          x$50 = $60M to 300M revenue

          But I believe Musk has projected 3% market penetration, which would be ~10M USA customers at ~12Mbps

          My point though was just that most of those customers would have to be rural. Each individual satellite will have to target a relatively small area of the surface. Even if they have enough flexibility for several to join forces to cover a city,you're still only talking about supporting thousands of customers out of millions of people living there - a tiny faction of a percent, compared to 3% overall.

          • (Score: 2) by deimtee on Thursday March 26 2020, @03:23PM (1 child)

            by deimtee (3272) on Thursday March 26 2020, @03:23PM (#975914) Journal

            Oops. Yeah, I cribbed 20Gps total from your earlier post and then used takyons 1Gps / connection.

            My $150M was per year. X 10(custs) / 12 months = $125M p.m. Brings it back to the same ballpark as yours.
            I've also seen a projected user price of $80/month which is 60% more. Given the uncertainties your range looks pretty good.

            The 10M USA customers might be after he gets all 42,000 satellites up. He isn't going to match that in AU, but if he gets permission to operate then half a million AU customers wouldn't be an unreasonable goal. Still a nice little earner for satellites that would otherwise be idle at that time.

            I would expect that small communities will band together and get one or two links, and then wifi everyone in range. There are places still on dial-up here, and a lot more on 2M/250K ~ 5M/500k ISDN or laggy GEO satellite. They could share 100Mbs between 10 houses and still be very happy.

            I really hope that they aren't location locked. There are a lot of reasonably well off "grey nomads" in Oz, retired people that live in caravans and head north for winter and back to the south for summer. A pizza box antenna and a tracking mount is well within their budgets. It sounds like a fun lifestyle.

            --
            No problem is insoluble, but at Ksp = 2.943×10−25 Mercury Sulphide comes close.
            • (Score: 2) by takyon on Thursday March 26 2020, @03:53PM

              by takyon (881) <takyonNO@SPAMsoylentnews.org> on Thursday March 26 2020, @03:53PM (#975927) Journal

              Oops. Yeah, I cribbed 20Gps total from your earlier post and then used takyons 1Gps / connection.

              [...] He isn't going to match that in AU, but if he gets permission to operate then half a million AU customers wouldn't be an unreasonable goal.

              Here is a source for 1 Gbps speculation:

              Elon Musk's SpaceX clears first hurdle to Australian broadband market [theguardian.com]

              Much remains a mystery about what Starlink’s internet services will be like in reality. In a November 2016 filing [fastcompany.com] with the US federal communications commission (FCC), SpaceX said it would be able to offer speeds of up to 1 gigabit per second for users, at “low cost”.

              1 Gbps is a magic amount that is impressive (which is why Google Fiber went for it) but can still be handled by most consumer equipment (maybe we'll see in uptick in 2.5 Gbps fiber soon?). But clearly a lot of people could get by with 100 Mbps, and would like to save a few bucks if they can.

              Australia's telecommunications regulator gave initial approval for SpaceX to operate its Starlink satellite network in the country. [cnbc.com]

              Australia could be one of the first major countries after the U.S. to get service, although the article notes some problems.

              I really hope that they aren't location locked. There are a lot of reasonably well off "grey nomads" in Oz, retired people that live in caravans and head north for winter and back to the south for summer. A pizza box antenna and a tracking mount is well within their budgets. It sounds like a fun lifestyle.

              The ability to get as much as 1 Gbps connectivity while living almost anywhere on the planet (outside of the extreme latitudes) or camping out somewhere will make those lifestyles more accessible and attractive to people. Get ready for the articles about literal "Digital Nomads" (with yurts). It may halt the outflux of young people from less populated areas.

              Des Moines is 'flyover country' no more: Millennials choose Heartland [desmoinesregister.com]

              --
              [SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
          • (Score: 2) by takyon on Thursday March 26 2020, @03:30PM (5 children)

            by takyon (881) <takyonNO@SPAMsoylentnews.org> on Thursday March 26 2020, @03:30PM (#975919) Journal

            Military is also a probable premium customer.

            SpaceX sees U.S. Army as possible customer for Starlink and Starship [spacenews.com]

            Air Force enthusiastic about commercial LEO broadband after successful tests [spacenews.com]

            A program known as Defense Experimentation Using the Commercial Space Internet, or DEUCSI, recently tried out SpaceX’s Starlink satellite broadband services and demonstrated download speeds of 610 megabits per second into the cockpit of a C-12J Huron twin-engine turboprop aircraft.

            --
            [SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
            • (Score: 2) by Immerman on Friday March 27 2020, @12:48AM (4 children)

              by Immerman (3985) on Friday March 27 2020, @12:48AM (#976149)

              Heck yeah they would be. Especially since we tend to fight our proxy-wars in out-of-the-way places that aren't easily connected to high-bandwidth, low-latency connections back to the drone-operators at home.

              Of course you probably wouldn't want to rely on a satellite signal on the battlefield proper (though it might be option #1 with a fallback). The transmission powers are low and easily jammed. But deliver that bandwidth to your base of operations and you can handle the last miles as appropriate.

              • (Score: 2) by deimtee on Friday March 27 2020, @01:05AM (3 children)

                by deimtee (3272) on Friday March 27 2020, @01:05AM (#976153) Journal

                It mightn't be that easy to jam. The ground station is highly directional and looking roughly straight up. The satellites are 500km away and moving fast. Swamping the signal might be doable, but then the ground station might simply switch to the next satellite if it loses the connection.
                Not to mention anything putting out that much interference on a battleground is radiating a "bomb here" message.

                --
                No problem is insoluble, but at Ksp = 2.943×10−25 Mercury Sulphide comes close.
                • (Score: 2) by Immerman on Friday March 27 2020, @03:23AM (2 children)

                  by Immerman (3985) on Friday March 27 2020, @03:23AM (#976201)

                  Bet you a $50 commercial drone with a $10 transmitter could do the job quite nicely over an impressive area. What's it cost you to shoot down a dispersed $1000 swarm worth of softball-size objects?

                  • (Score: 2) by deimtee on Friday March 27 2020, @12:30PM (1 child)

                    by deimtee (3272) on Friday March 27 2020, @12:30PM (#976263) Journal

                    No bet. :)

                    Doesn't mean that the military won't pay a shit load to use it in any area that isn't an actual battlefield.

                    --
                    No problem is insoluble, but at Ksp = 2.943×10−25 Mercury Sulphide comes close.