Follow Slashdot blog updates by subscribing to our blog RSS feed

 



Forgot your password?
typodupeerror
×
Communications Network The Internet United States

Can SpaceX's Starlink Broadband System Deliver Less Than 100ms Latency? (arstechnica.com) 143

Proudrooster writes: Can SpaceX's Starlink deliver less than 100ms latency? That is the $16 billion dollar question as the FCC looks to pump more money into rural broadband (PDF). Will rural America ever get broadboand? What will America get for giving billions of dollars to incumbent providers? Is this all just FUD or a publicity stunt by the FCC to stop states from building out their own networks? Will American's continue to pay the highest prices on earth for broadband? Will broadband and Internet access ever be considered a human right. So many questions, so few answers. Stay tuned. "FCC Chairman Ajit Pai backed off a plan that would have completely prevented SpaceX and other LEO companies from applying for rural-broadband funding as low-latency providers," reports Ars Technica. "But the FCC's full order was released today and suggests that SpaceX will have a tough time convincing the commission that its service will deliver latencies below the FCC standard of 100ms."

The FCC has "serious doubts that any low-Earth orbit networks will be able to meet the short-form application requirements for bidding in the low-latency tier" and that companies like SpaceX thus face a high chance of being rejected when they apply for funding as low-latency providers: "Service providers that intend to use low-Earth orbit satellites claim that the latency of their technology is "dictated by the laws of physics" due to the altitude of the satellite's orbit. We remain skeptical that the altitude of a satellite's orbit is the sole determinant of a satellite applicant's ability to meet the Commission's low-latency performance requirements. As commenters have explained, the latency experienced by customers of a specific technology is not merely a matter of the physics of one link in the transmission. Propagation delay in a satellite network does not alone account for latency in other parts of the network such as processing, routing, and transporting traffic to its destination. Short-form applicants seeking to bid as a low-latency provider using low-Earth orbit satellite networks will face a substantial challenge demonstrating to Commission staff that their networks can deliver real-world performance to consumers below the Commission's 100ms low-latency threshold."

As Proudrooster mentioned, if SpaceX is rejected from the low-latency category, it "will be at a disadvantage in a reverse auction that will distribute $16 billion -- $1.6 billion yearly, over ten years -- from the Rural Digital Opportunity Fund (RDOF)," the report says. "The FCC will prioritize low-latency networks when awarding funding, so SpaceX and other LEO providers could come up short against terrestrial networks. Even DSL providers would have an advantage over LEO networks in funding battles if the satellite companies are placed in the FCC's high-latency category."
This discussion has been archived. No new comments can be posted.

Can SpaceX's Starlink Broadband System Deliver Less Than 100ms Latency?

Comments Filter:
  • by backslashdot ( 95548 ) on Saturday June 13, 2020 @02:12AM (#60178384)

    If they are in LEO, theoretically they should be have the same or only neglibily higher latency than fiber. 400 miles up + 1000 miles straight (satellite to satellite) + 400 miles back down. That has a theoretical low of 10 milliseconds.

    • by Rei ( 128717 ) on Saturday June 13, 2020 @02:49AM (#60178426) Homepage

      FCC's argument is not that the satellites would have low latency - they acknowledge that that would be a low-latency leg (the speed of light in fibre is 2/3rds that of light in a vacuum) - but that the ground stations might add too much latency to be considered "low latency".

      I don't get this argument. The ground stations are intended to plug as closely as possible directly into backbones. Rarely do I find high latency at backbones; for me, whenever there's latency problems, it's at the end of the journey, either routing through (many) hops to get to a backbone from home, or the same on the other side.

      Another source of latency for my ground-based internet service is data going the wrong way. For example, when I geotraceroute to Slashdot, my data is going from west Iceland, to east Iceland (the long way around), to London, to DC, to Colorado, to Arkansas, to California. Lots of "going the wrong direction" there. But Starlink should always drop you onto a backbone that's "in the right direction" and "going the right direction". And any satellite-satellite communications should also be "in the right direction".

      Once you're on a backbone, it's no longer up to Starlink, but at least in terms of getting there, I can't see how Starlink would be the problem.

      • by Rei ( 128717 ) on Saturday June 13, 2020 @03:01AM (#60178436) Homepage

        I'd personally love it, but the website for the service went live yesterday [starlink.com], and for the first year they'll only be accepting people from North America. The currently launched satellites of course fly over Europe, but you have to build out ground stations before you can realistically use the network in a given area. Otherwise, for someone in Europe to, say, access their local bank, satellites would have to bounce all traffic across multiple satellites to ground stations in the US and Canada, get on a backbone there, then come back across the Atlantic :P And for Iceland it's right-out, since the satellites launched thusfar don't go this high latitude.

        It's not a problem for someone in the US. Someone in, say, Podunk, West Texas wants to access a website in Iceland, the signal goes up and laterally to the northeast (most likely 1-2 bounces), drops into a US backbone somewhere between Dallas and NYC, and then heads across the Atlantic as usual. But in the reverse scenario...

      • by AmiMoJo ( 196126 ) on Saturday June 13, 2020 @03:49AM (#60178494) Homepage Journal

        Not so much the ground stations but the shared nature of any radio frequency network. The more users per satellite the fewer communication slots available per user, so the longer a packet has to wait for a slot before being transmitted.

        That's why latency on mobile networks is so horrible - not only is it high but it fluctuates wildly as other users request slots or move between base stations. I did a lot of work on this a few years back for time sync and for example NTP doesn't work very well because the latency varies between sending a request and getting the response so much.

        We will have to wait until we see some real world stats in areas with a decent number of subscribers.

        • Why do you think starlink wants 12000 plus satellites?
          Starlink already has more satellites in orbit than any one company. With over 500 and plans to launch another 150+ by the end of June

          Every starlink launch is another 50-60 sattelites

          By having somany targets above you gets lots of options.

          • by Rei ( 128717 ) on Saturday June 13, 2020 @06:04AM (#60178640) Homepage

            They just launched another 58 a couple hours ago. The cadence is really impressive.

            • by rew ( 6140 )

              Watched the webcast.... "Next batch, later this month". There is about 2 weeks remaining in this month....

              Access to "orbit" has moved from: you wait years for a slot on a rocket, to just say when you want to launch.

              The cost per kg is about 2200 dollars. That's list price. This means that a small 1kg cubesat can go up for "doable for a private person" prices.... The only thing is they rideshare in 450kg chunks at the moment. Give it a couple of years and that will change.

          • by AmiMoJo ( 196126 )

            It's hard to calculate the area each satellite would need to cover because there needs to be overlap of the circular footprints. They are going to be massive though. The Earth's surface is 510 million square kilometres.

        • by Rei ( 128717 )

          The more users per satellite the fewer communication slots available per user, so the longer a packet has to wait for a slot before being transmitted.

          Sounds like you're talking about bandwidth limitations rather than latency limitations. If you're not at bandwidth limits, you're not stuck in a queue.

          Obviously, to avoid bandwidth limitations, Starlink will need to limit its consumer base in densely populated areas, where "densely populated" is primarily relative to the size of a spot beam of the lowest-orbi

          • Re: (Score:3, Informative)

            by AmiMoJo ( 196126 )

            Latency increases with link saturation. It doesn't have to be anywhere near saturation to go up, in fact it will go up with even low levels of traffic. It's not only a function of bandwidth either, it depends on the number of users. Two high bandwidth users will see less latency than 200 low bandwidth users.

            This is a very well understood and common problem with things like mobile networks and wifi.

            • by robbak ( 775424 )

              And that depends on how well the Quality-Of-Service software on the terminals and satellites works. If you always allow low-latency service traffic through first, you can have high throughput and low latency at the same time. Indeed, if you avoid buffer bloat - keeping queues short and dropping packets fast instead of having huge buffers and timing packets out of them after seconds of waiting - the standard internet protocols sort out much of that.

              • by AmiMoJo ( 196126 )

                It's hard to do good QoS on packets these days because everything gets encrypted. Not just VPNs, a lot of traffic is obfuscated to avoid firewall problems by making it look like something else.

                • by zekica ( 1953180 ) on Saturday June 13, 2020 @08:37AM (#60178880)
                  Take a look at FQ-Codel/Cake and FQ-PIE to see how you can do good QoS without looking at the contents of the packages. Also, if they use OFDMA they can shrink the bandwidth per user but no user will have to wait for their timeslot as they will all be simultaneous. With even 1Mbps, latencies of 20ms are achiveable and with 10Mbps they can reduce them to 2ms (excluding propagation and forwarding time - which can be as low as 10ms.

                  So in total, latencies of 20ms are easily achivable if you don't make the same mistakes as LTE.
            • That's technically right, but irrelevant in practice. If you send 20 packets a second, you can expect no more than 50ms latency from queuing, if access is fair, the link is not oversubscribed and there are no large buffers (i.e. the network operator does a good job). 20 packets a second is 30kB/s. At 1Mbps, latency from queuing is less than 12ms. It doesn't matter if the link is saturated by 2 high bandwidth users or 200 low bandwidth users. The high bandwidth users will see lower latency simply due to thei
              • by AmiMoJo ( 196126 )

                Any counter argument needs to account for why mobile networks and WiFi are so bad. Can you explain that?

                If there was a fix then someone somewhere would be able to demonstrate it on their network.

                • Wifi uses frequency bands which are also used by lots of other protocols and other networks. The reason it uses the 2.4GHz band in the first place is that it's an unlicensed band and it's unlicensed because it's where lots of noise is. The Wifi protocol isn't designed to minimize latency because that would be pointless on such an unreliable channel. As for mobile, LTE does have very low latency. Wifi and mobile aren't the only shared medium networks. Cable shares frequency bands among many users too and you
            • This problem can be "solved" already by baking latency into the system and giving guaranteed time slices. This doesn't eliminate latency, but it does solve the jitter problem.

              • Interesting, sounds similar to how compilers will pad a memory allocation to make it DWORD or WORD boundary aligned, speeding up loading into CPU registers. Wastes small bits of memory but does make overall processing faster since you want to load on those boundaries anyhow.
              • Like GPON. But GPON has the benefit of substantial bandwidth over fiber available.

        • We will have to wait until we see some real world stats in areas with a decent number of subscribers.

          Couldn't they just model it? It seems to me they already have both specs and theory that would allow for fairly accurate average-case and worst-case latency predictions.

          • by cusco ( 717999 )

            Conservatives don't accept models as evidence, because they're all science-y. If you start to believe in models then you start to accept horrors like global warming and an Earth more than 6,000 years old, and we can't have that.

        • Jitter (change in latency) seems to be endemic to wireless networks but latency itself will tend to be a function of the network's design. 10,000 TDM timeslots in a cell will show high latency waiting for a slot but TDM isn't the only answer and even with satellite, cells don't have to be 10,000 users wide. Especially in flight over sparsely populated rural areas.

      • by ceoyoyo ( 59147 )

        The FCC statement seems pretty weird. StarLink pays a very small price to go straight up, but makes it up in faster transmission time. It probably also has fewer hops from satellite to satellite than a typical broadband provider would have going through the switching equipment at the end of e.g. a rural road, probably a regional one, and a couple in the nearest city before you get onto an actual backbone.

        It also shouldn't be too hard for SpaceX to demonstrate the latency of their system. The other LEO provi

    • Well, that is just the backhaul. It goes up to orbit, then down to some earthstation located at a major internet backbone (I assume), then there will be additional latency from backbone to final destination. But yeah. Being under 100ms seems very possible. Even if not 100% of the time, much of the time. Routers are designed to dump packets as fast as they can, not store them.

      • I think some have forgotten there's a long history of satellite communications from ships, and planes. So people know how to communicate with something in the middle of nowhere and moving fast. The primary difference is making all that affordable.

        • by robbak ( 775424 )

          Internet on ships has been horrible for ages. Mostly, it was geostationary satellite internet, and that with smaller, steerable dishes that can't collect enough signal to mailtain a high speed link. In recent years you have had iridium, but their initial satellites focused on voice - it is only the new collection that can do reasonable data - but never enough to provide a whole ship with service.

          If SpaceX can provide ships with solid gigabit service, it will make many people very happy.

          • Not only ships, couldn't this also equally apply to planes? That will make even more people happy. But more to your overall point, I agree, and commercial shipping will benefit hugely from having solid reliable high speed datalinks, to orchestrate traffic, weather conditions and deliveries worldwide. I imagine Maresk would love to have this. Basically the laziness of internet connected manufacturing, for global shipping at sea.
    • by DrXym ( 126579 )
      Tracking & handshaking with satellites, dropped packets, congestion, hop times (from the data center) should also factor into it. And since the satellites are constantly in motion, they aren't always directly up and there may also be handover delays as you switch from one satellite to another.
    • Considering the steps for the fiber rollout were:

      DO

      Promise to make fiber available in a region of the country

      Take the subsidy for making fiber available

      LOOP

      It shouldn't be hard to beat. Actually requiring SpaceX to follow through would be quite a shift from how this game has been played for the past 30 years.

    • by DRJlaw ( 946416 )

      What is this nonsense?

      Title: "Should be faster than fiber"
      First sentence: "If they are in LEO, theoretically they should be have the same or only neglibily [sic] higher latency than fiber."

      Does not compute.

      400 miles up + 1000 miles straight (satellite to satellite) + 400 miles back down. That has a theoretical low of 10 milliseconds.

      With instantaneous reception, processing, and retransmission from both satellites and a ground station?

      Does not compute.

      It would help if you'd actually consider all of the pote

      • I agree... the GP comment is presuming a perfect world and perfect transmission/reception/retransmission. I think what they fail to realize is that there is no such thing as a perfect world, there will always be retransmits and there is packet switching delay ALWAYS in a network. In a perfect world, fiber provides light speed communication at virtually zero latency worldwide... doesn't mean that's the way it literally works.

        In order to calculate the real world latency you need to know a lot of information t

      • by lgw ( 121541 )

        FCC was claiming that the laws of physics forced latency higher than 100 ms. That is obviously wrong. The laws of physics are relevant to the length of the signal path and the speed of light in the relevant medium.

        Switching and queuing delays are the same problem for both sat and ground networks, and not a problem with "the laws of physics" (at least, not at the scale of 100 ms). The FCC is obviously confused about how much closer these sats are then GEO. Their statement is true of GEO, false of LEO. S

        • by DRJlaw ( 946416 )

          FCC was claiming that the laws of physics forced latency higher than 100 ms.

          It wasn't, and I even quoted the FCC statement in the very message that you replied to, you moron. Here it is again:

          As commenters have explained, the latency experienced by customers of a specific technology is not merely a matter of the physics of one link in the transmission. Propagation delay in a satellite network does not alone account for latency in other parts of the network such as processing, routing, and transporting traf

          • That is the revised language. Initially the FCC excluded all satellite bids because it was 'impossible'. Aka geo sats are physically impossible therefore all sats are impossible.

            The new framework is sensible. "total system latency demonstrated" vs a blanket exclusion.

            Starlink and terrestrial point to point microwave should be treated equally because you can create a crap point to point WAN as well. I've been on crap terrestrial systems as well and yet they were allowed to bid in the low latency.

            The only rea

            • by DRJlaw ( 946416 )

              That is the revised language. Initially the FCC excluded all satellite bids because it was 'impossible'. Aka geo sats are physically impossible therefore all sats are impossible.

              Bullshit. Provide an FCC link with an actual quote.

              You'll only find this [fcc.gov], where the initial language does not substantially differ. It says:

              108. While satellites in low earth orbit may not be subject to the same absolute physical
              latency limitations as higher-orbiting satellites, we disagree that the altitude of a satellite's orbi

  • Not true (Score:5, Informative)

    by kubajz ( 964091 ) on Saturday June 13, 2020 @02:13AM (#60178386)
    From TFA: USA has 9th most expensive broadband from the 34 surveyed OECD countries. Summary goes for sensational "most expensive in the world". Shame.
    • by aliquis ( 678370 )

      Exactly.

      Here in Sweden back in 2001 I paid 200 SEK for 10 mbps and now in 2020 I pay 399 SEK for 200 mbps from the same company. I could get a bit lower price easily though. As for the speed difference it's just upgraded to that it's not like I've picked a faster package.

      If you compare Sweden to the USA back then 10-20 years ago what USA offered seemed bad but it's my impression nowadays that the situation have improved and that the difference is no longer all that big. Then again there may still be a bunch

      • The problem with the USA is that we're highly spread out. The other problem is that we gave the telcos billions of dollars to build out the last mile, with no accountability. So they spent the money on executive compensation, stock dividends, stock buybacks and so on, and didn't build out the last mile at all in most locations.

        Another problem is that in many locations the cable company has a monopoly on the right of way granted by local government, usually again without any accountability. They got the mono

        • My in-laws in Sweden now have gigabit fiber to their summer house in the woods in a community of ~1,000 covering ~10 square miles. It is a recent thing; they were previously on a 10Mb wireless link, but the point is Sweden is still making progress rather than complaining that things are too spread out.

          And to the GP: you can get service as slow as 200Mb in Sweden?!

  • Try it and see? (Score:4, Insightful)

    by Gabest ( 852807 ) on Saturday June 13, 2020 @02:37AM (#60178408)

    Am I crazy to suggest that as an answer? I thought there were already in the testing phase, with working equipment.

    • Testing yes, and trying to scale it up. A launch here in about 2 hours will be their 9th Starlink launch, and at 60 satellites per Falcon rocket, there are only about 500 satellites, out of a proposed constellation of 30K or more. It'll take a while.

    • Am I crazy to suggest that as an answer?

      I am open to trying it, if given the chance. I live in the northern US, and would love to have a viable alternative to Comcast.

      So I signed up for notifications - we'll see if any concrete information comes from it. They seem to be hiding any specifics about the service, though, which doesn't exactly fill me with confidence.

    • To prove any test with a limited load will approximate a realistic case would require them to expose a lot of trade secrets.

      Which I don't think they'll be able to get out from under. They'll have to expose intimate details about tracking, handover, relaying, ground stations etc etc so independent third parties can run simulations (and competitors can accurately calculate their costs). It's trade secrets or subsidy it seems.

  • Just wait until Elon puts servers in orbit.

    Exactly like the terrestrial Internet depends on CDNs to deliver services and results in a timely manner, so will the Orbital Internet(TM)

    It's a logical evolution

    • Caching servers maybe. Though realistically just putting those in the building with each ground station is almost as good.

    • by dwywit ( 1109409 )

      How would orbital servers deal with heat?

      Bloody big radiator fins on the shadow side?

    • Orbital load balancers. LOL.

    • by robbak ( 775424 )

      Latency between a server and a user isn't really the issue. No one cares if it takes 50 or 500ms to set up a download - what counts is how fast it comes when it does.

      Latency matters when it is between users - packets between two users video conference screens. Servers in orbit won't help that.

      • Except I see 20-30 connections for each web page I load. Heck, noscript is showing 13 different connections for this page. We are not talking about an ftp style single server/client. Using the web today is a massive number of connections.
      • by ceoyoyo ( 59147 )

        Lots of people care quite a bit about latency to servers. These people are often called "traders" and "gamers" although the latter label is probably sufficient for both.

    • "Ya sure, I've got about 5 years experience with this technology"
  • For example since the satellites movie relative to the ground you need to track them. This is not cheap no matter if you do this by physically moving your dishes or using beam forming. I doubt this would be a feasible alternative for individuals.

    So what's left is the backbone. For that you are faced with the alternatives of having a simple fibre in the ground, which is cheap in rural areas and has very low operating costs, or a fairly expensive ground station with high operating costs.

    In any case you'll sti

    • I would think a technical forum would know that "cheap" is a rather relative target. Not just in the demographic sense, but the moving pace of technology sense. It's why most of you are typing responses on an "affordable" but powerful computer. Same applies to base stations even if it's much bigger.

    • Fiber in the ground is cheap in flat rural farming areas. But there are also mountainous rural areas where both fiber and wireless ISP's have challenges getting to all customers.

    • by robbak ( 775424 )

      That is being solved. Yes, they will be using beam forming - and doing it cheaply by mass producing the antenna panels. Beam forming hardware is expensive because it is made in small amounts using labor-intensive processes. They will be targeting a similar terminal cost to current terrestrial fixed wireless terminals.

      Forget niches - Stalink's market is every place where fiber-to-the-premises isn't feasible. That's everywhere outside of cities.

    • Yeah, Fiber is "cheap", right. Do you have any idea how much it is to install a tap and run fiber? At all?

      My friend lives just outside a town. The fiber lines run on poles in front of his house. If he wants fiber a tap to the line and running the line the 80 meters to his house would cost almost $4,000. That isn't "cheap" by any stretch of the imagination. That's also the cheapest quote, not something you want to just grab without verifying the quality of the company.

      • by ceoyoyo ( 59147 )

        In terms of infrastructure, that's pretty cheap. Compare to what it would cost to run a water pipe, or road.

    • by cusco ( 717999 )

      In the ground? Probably not. The reason why you see miles of telephone and electrical cables on poles throughout the rural areas of the world is because trenching and laying underground cable is slow and expensive. A crew of half a dozen guys might be able to trench half a kilometer on a really good day in a flat place with loose soil and nothing else already in the ground that they need to avoid. That same crew could tack a couple of kilometers of cable to existing poles. Plus above ground cable is a

    • by ceoyoyo ( 59147 )

      You can buy a wifi access point for $50 that does beamforming.

      • Yeah, but not at those frequencies and only with a hand full of antennas. That beamforming is far to coarse for such an application.

        • by ceoyoyo ( 59147 )

          It's not the 1990s. You can make a phased array antenna on your desk or in your garage. If you can figure out how to use Eagle or similar software you can design one on a PCB and send it off to China to be manufactured for you for a buck.

    • by lgw ( 121541 )

      For example since the satellites movie relative to the ground you need to track them. This is not cheap no matter if you do this by physically moving your dishes or using beam forming. I doubt this would be a feasible alternative for individuals.

      What on earth are you talking about? Obviously the engineers at SpaceX solved this problem early on, or they wouldn't have launched hundreds of sats. It's a pizza-box sized ground station that just needs to face the sky, from what we know. What "operating costs"? What "distribute the network to the residents"?

      Are you talking about something other than Starlink? I'm seriously confused here.

  • Need Other Providers (Score:5, Informative)

    by kenwd0elq ( 985465 ) <kenwd0elq@engineer.com> on Saturday June 13, 2020 @02:52AM (#60178428)

    We need to get some new players into the arena of broadband service. For example, I was researching buying a home in Bandera, TX, and discovered that the small, local, electric utility Bandera Electric Cooperative not only delivers power, but also gigabit fiber in their service area. And the prices for gigabit service with no caps was REMARKABLY reasonable; low, even. Their area is more rural than suburban, but even in the sticks, somebody is delivering AC power - and if they own the poles, there's no reason why the shouldn't run fiber as well.

    https://www.banderaelectric.co... [banderaelectric.com]

    I ended up buying a house closer in to San Antonio, but network connectivity wasn't the reason.

    • Power utilities are kind of a sleeper candidate here, for reasons mentioned, and it fits into the "smart grid" initiative. The incumbents are blocking the traditional broadband incumbents, but not electrical utilities.

      • My local utility [wikipedia.org] did this back in the 1990s. They figured out that by running fiber around town, they could monitor electrical loads very precisely, and this would allow them to save tons of money on peak power purchases from the grid. And as a side-bonus, the fiber would also act as a backbone for cable TV and broadband internet. They proposed a bond issue (or maybe it was a local option sales tax, don't remember exactly) and put it up for a vote. It passed, of course, and ended up paying itself off (in po

  • Does this have to be asked on every technical forum?

    Yes. 8ms to LEO, 8ms back, 8ms to switch the packet if you're using old hardware, and another 8ms if you're sending it off to another bird.

    That's 24ms ground station to ground station.

    Could we POSSIBLY do a Google search before again ask.... oh.... it's a weekend... we have to have stupid slashdot stuff. I guess that makes sense. Please post more questions like "Can SpaceX REALLY deliver two men to the ISS?" or "Is it possible the sky will not be blue

    • Where are you getting that 8 ms one way timing? At an altitude of 550 km, that's less that 2 ms to LEO. So even if I assume your other 8 ms timings are correct, I've got to cut your "24 ms" down to 12 ms. Next time, do the math.

    • by cusco ( 717999 )

      I live in Seattle, you inconsiderate clod! The sky will probably be grey fading to whitish patches all weekend.

  • Starlink is supposed to be at 550 km. Geostationary orbit is 35,786 km. Even if starlink can't hit 100ms figure, it will be way, way better than Hughes and all the geostationary satellite ISP's. I love to hate on Elon about a lot of things, but starlink, if it works reliably, will fill a very, very much needed market segment. There are a huge number of under-served customers right here in the US, not to mention the whole world (if the price is right).

    Routers generally try to avoid storing more data than nec

    • I am not sure the FCC should give ANY companies money, but if they give it to anyone they should consider giving it to SpaceX for sure.

      Shouldn't, but they've been flushing money down the incumbent hole for years and little to show for it. So maybe Ajit Pai could be a little less transparent about where his loyalties lie and start taking some risk on an up and coming. Worse comes, the US owns a new constellation. Same can't be said for the incumbents.

      • I am not sure the FCC should give ANY companies money, but if they give it to anyone they should consider giving it to SpaceX for sure.

        Shouldn't, but they've been flushing money down the incumbent hole for years and little to show for it. So maybe Ajit Pai could be a little less transparent about where his loyalties lie and start taking some risk on an up and coming. Worse comes, the US owns a new constellation. Same can't be said for the incumbents.

        Have to agree with you. I have faith that Musk is actually trying to solve the problem, whereas the others are just trying to harvest taxpayer money and they have probably figured out by now that they can keep getting more as long as the problem persists. So they actually have no incentive to fix it. At least not until we have a different chair.

      • by cusco ( 717999 )

        Risk? You must be kidding. Pai is going to exit this farce of "public service" and step directly into lucrative board positions with the companies that he's been granting favors to for his entire term, the same as Michael Powell. Musk isn't going to give him free stock and a no-work income stream so he has absolutely no incentive to do anything but put roadblocks in Starlink's path.

  • by Vitus Wagner ( 5911 ) <vitus@wagner.pp.ru> on Saturday June 13, 2020 @03:55AM (#60178504) Homepage Journal

    > Will American's continue to pay the highest prices on earth for broadband?

    Of course. Because they have that money for broadband providers to rob from them.
    Prices can drop only if consumers are not willing to pay. But people really need broadband, and market is so big, that any crime would be commited to stop competitions. I've read numerous stories how commercial provider sue municipalities to stop providing cheap connectiviy. Now they lobby FCC to kick low-satellite companies from auction.

    • Not surprising since the robber baron days that companies do things at the expense of others, but what's different is they're just so transparent about it. It's like robbing the till in front of a whole crowd, and saying, eh, so what?

  • by Vermonter ( 2683811 ) on Saturday June 13, 2020 @06:02AM (#60178636)

    This whole rural broadband grant was designed to be a gift for comcast and other major us ISP's who don't actually invest to upgrade their networks. Everything in the system and rules from the FCC was designed to favor them. They did not foresee someone actually managing to build any sort of competition, and now Pai is trying to put out that fire.

  • Missed The Point (Score:5, Insightful)

    by ytene ( 4376651 ) on Saturday June 13, 2020 @06:24AM (#60178660)
    Top tip: whenever you see "FCC" in a document or statement that expresses a view or opinion, simply translate the above acronym to,

    "The cartel of telecommunications companies that hide behind the FCC"

    and you will get a clearer understanding of what is really being stated.

    In this specific case, what we're seeing is that SpaceX are launching more and more of their Starlink satellites, which means that the threat they represent to the telecommunications cartel is growing steadily. Add to this the fact that the telecoms companies are telling the FCC that they need *billions* to run out local loop connections to rural properties and communities and you can see how Starlink has quickly become a significant and urgent threat to the "easy money" that the telco's have been squeezing out of the FCC for years.

    So it's entirely in keeping with those companies to continue to use their corrupt proxy, sorry, the FCC, to ensure that Starlink can never become a threat to their profits.
    • by DRJlaw ( 946416 )

      Top tip: [ad hominem manifesto omitted]

      There's a very simple answer to this. SpaceX needs to publish its own data concerning its latencies and permit independent testing. It has 539 satellites in orbit but hasn't begun to offer any service, and hasn't published any data. Its applications are due by July 15th. Promises that it will meet the requirements sometime after the application is due won't cut it.

      Time to show your cards, Elon.

      • by cusco ( 717999 )

        Well, if they required that then Comcast and Verizon would also have to show their miserable latency to outlying areas, and they really don't want to do that. They might have to spend some money upgrading the rural link routers, which would affect next quarter's stock price.

        • by ytene ( 4376651 )
          Not only are you right in pointing this out, but there are a couple of additional things to consider here...

          Firstly, from a technical perspective, whilst latency is important, it isn't everything. The FCC's request would seem to suggest that they believe that the round trip distances involved in getting signals up to LEO and back down again will have a material impact on latency. Obviously, since there are distances involved, they are of course right. But how much of a factor is that? Really?

          Consider
  • So he can run an actual free anonymous Internet by his satellite and be capable enough to defend the freedom of humanity and his satellites?

    • I think you missed the part about the annual 1.6B from the government that he will be getting a slice of. Don't bite the hand that feeds you.
  • This whole process is just another example of government giving the public's money to corporations, so the corporations can develop infrastructure which they will then own, and on which they will continue to make substantial profit from the same public for decades.

    How about a system wherein the various jurisdictions involved are shareholders in the enterprise? That would allow at least some of the profits to revert to the people on whose backs the scheme is being funded.

  • I think there are a lot of practical issues beyond the theoretical latency, packet loss, jitter, link saturation, queuing problems, etc. The Elon Musk halo and adding 'in low earth orbit' does not solve well known network issues.

  • Elon stated a couple months back that private beta testing was coming in a month or so, with public beta in four.
    No point in trying to convince with argument the FCC when you can point to delivered latency.
  • Wife and I are both on the fast track for telework. Both of us were born and raised in the south bay, and we're dismayed at what a shithole it's become. Sky high taxes, crime, poverty everywhere. California as a whole has become a bad value in living. We just took a roadtrip through Oregon (and returned yesterday) We drove the entire state up the 101 from Brookings to Astoria, and found a ton of places we would love to live. We were amazed at the low cost of everything, how clean it is, how we only sa

"The vast majority of successful major crimes against property are perpetrated by individuals abusing positions of trust." -- Lawrence Dalzell

Working...