Stories
Slash Boxes
Comments

News for nerds, stuff that matters

Slashdot Log In

Log In

Create Account  |  Retrieve Password

Japan Launches "Super-Speed" Internet Satellite

Posted by kdawson on Sat Feb 23, 2008 04:20 PM
from the fiber-in-the-sky dept.
A number of readers wrote in about the launch this morning of a Japanese H-2A rocket carrying a Kizuna ("Winds") satellite into orbit. Kizuna is intended to provide "super high-speed data transmission" for Japan and Southeast Asia. The news stories on the launch, such as the AP's linked here, are short on technical detail. For example they say the satellite successfully achieved orbit 175 miles above the earth — hardly suitable for Internet communications to a specific area on the surface (remember Teledesic?). Reader nebulus4 provided a link to the Japan Aerospace Exploration Agency site with an illustration and a little more detail. Such as the fact that Kizuna is destined for geosync orbit, and that a 45-cm antenna will equip eventual users for 155 Mbps down / 6 Mbps up, whereas a 5-m antenna will allow enterprises and ISPs to tap into 1.2 Gbps down. Given the latency to geosync orbit, you probably wouldn't want to use Kizuna to play an online shooter.
+ -
story

Related Stories

This discussion has been archived. No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
 Full
 Abbreviated
 Hidden
More
Loading... please wait.
  • by Anonymous Coward on Saturday February 23 2008, @04:22PM (#22529322)
    ...the RIAA and MPAA today announced a plan to knock the satellite out of orbit with a missile to "protect the public".
  • by Zorbo88 (1240952) on Saturday February 23 2008, @04:28PM (#22529390) Homepage
    So a subsistence farmer in rural Indonesia gets a better download speed than me, a sophisticated suburban Australian. Awesome.
    • Re: (Score:3, Insightful)

      Well nothing really makes you better than him to begin with, so you're not generally entitled to better internet than him.
          • Re: (Score:3, Informative)

            Too bad you're wrong, Mr. Troll: those phrases haven't been around for "cednturies [sic]." The phrase third world [etymonline.com] was coined by Alfred Sauvy in the 1950s. He also retroactively coined the words first world and second world to apply to already existent categorical differences between the Democratic West and the Communist East. I believe the phrase you confused with first world is the Old World. However, the Old World does not merely apply to the wealthy European nations, but all of Eurasia and Africa
    • Re: (Score:2, Insightful)

      If he could pay for it, it's sorta like the T3 you can't buy because you don't have the cash either. At any rate, I'll wait to see how much they really get out of this, I've had a friend with satellite service that's way in the outbacks and it was expensive, unstable, underdelivered on bandwidth and latency was higher than advertisied and it was in general a pain to use. He jumped to cable first chance he got, I don't remember which but it was one of those bloody-sucking underdelivering monopolies that get
    • So a subsistence farmer in rural Indonesia gets a better download speed than me, a sophisticated suburban Australian. Awesome.


      Have you ever tried using satellite internet before? You'll soon realize that speed isn't everything.

      -matthew
    • Wait for the url filter to kick in.
      Then you will have subsistence internet too.
      • Re: (Score:3, Informative)

        You laugh about the Kiwis, but we get virtually all our internet here in Oz via the Souther Cross Cable [wikipedia.org] system. A system with NO australian ownership whatsoever. The majority owner? Telecom New Zealand, with a 50% stake.

        Yup, if it weren't for the Kiwis we'd still be sending our email by morse code. (The next biggest cable, between Australia and Japan isn't anywhere near big enough and came online several years after the SCC)

        Gotta hand it to them; they wanted a big cable for themselves but probably couldn't
  • by dsginter (104154) on Saturday February 23 2008, @04:30PM (#22529400)
    22,233 miles to the satellite
    round trip = times 4 = 88,932 miles

    speed of light (wave propagation) = 186,282 mi/sec

    latency = 88,932 / 186,282 = 0.477 seconds (on top of regular network latency)

    Curse you speed of light. You win again!
    • That just means that you need to have a large TCP window to compensate the large bandwidth-delay product. No real problem. The connection sucks for anything interactive, but bulk is just fine.
      • by Dunbal (464142) on Saturday February 23 2008, @04:36PM (#22529448)
        The connection sucks for anything interactive

              Except, possibly chess.
        • Re: (Score:3, Interesting)

          The connection sucks for anything interactive

          Except, possibly chess.

          I know you were joking, but as an administrator on a chess server, I can tell you that people get pretty pissed off when lagging half a second. It's acceptable for playing long games, but most over-the-net chess games are 1 to 5 minutes per player per game. Yes, it's a whole different game that just shares moving rules with "chess".

      • Re: (Score:3, Informative)

        That just means that you need to have a large TCP window to compensate the large bandwidth-delay product. No real problem. The connection sucks for anything interactive, but bulk is just fine.
        I've got satellite. Latency effects more than you think. Yes for big files its fine but p2p, web surfing, voip, if you voice chat. Sometimes my latency is 2-3000ms. id rather have a 384kb dsl line at home and just grab my .iso's at work.
        • Depending on the cost and throughput, i would consider getting satellite in addition to my DSL for bulk transfers...
          Use bittorrent to download to a fast server, and them download it over the satellite link from there.
    • Why not put up a bunch of them up (like GPS) and just hop around to available Sats as they come over the horizon ?

        Note:
      I know nothing at all about this stuff, so be nice when you call me a idiot.

        • Unless of course you sell service to people all over the globe with a constellation of LEO satellites.

            In any case, a Molniya orbit would only require three satellites for coverage, looks ideal for Japan as a nation, and the perigee can be as low as ~400km. The round-trip latency for 400 km would be (400*4/300,000), or 5ms (if my mental mathematics is not off by a decimal point or so).

          Yes, you'd need three satellites, admittedly.
          • a Molniya orbit would only require three satellites for coverage ... The round-trip latency for 400 km would be (400*4/300,000)

            Problem is, a Molniya orbit requires three satellites for coverage at the apogee, which is at about the same altitude as the geosynchronous orbit. At the perigee the satellites move faster, so you need more of them to keep one always on sight.
        • Not to mention that your dish has to 'track' the satellite to get a decent connection time. Then it has to swing back to the 'start' position to pick up the next satellite in orbit.
    • round trip = times 4 = 88,932 miles

      what, you don't think it would be appropriate to put your servers in geosynch orbit as well?
    • From what I've read latency over 1 second is typical. It makes sense; the data is probably spread out and repeated so that the lost parts of the signal can be pieced back together. I mean, phone modems encode data like this these days too...
      • Is there a reason the satellite has to be up this high? Could it be at a lower altitude?

        Not if the satellite is going to appear to sit in one spot in the sky. At geosynch altitude, the sat has an orbit of 24 hours, thus, appears to be stationary.

  • 175 miles (Score:5, Informative)

    by heroine (1220) on Saturday February 23 2008, @05:51PM (#22529932) Homepage
    175 miles is the separation altitude for the rocket. Satellites usually boost themselves to geostationary orbit. The Delta IV heavy can blast all the way to geostationary orbit but no-one can afford it.

  • IT SUCKS!!! I install WildBlue, 1 of the 2 main Sat. ISP's. It doesn't matter how fast the connection speed is, the latency SUCKs. It averages 1200-1800 ms (no that's not a typo, check it out if you want). You can not play online games, outside of backgammon. The only thing I can say for it is, it is better than dial-up, although you can play some online games w/dial-up.
  • Any Idea how the uplink is supposed to work? Can thousands of 45cm dish all communicate with the satellite at 36,000km simultaneously? I know that years ago you needed a DSL/Dial-up connection for the uplink, is the not the case anymore?
  • is something to solve the last "175 mile problem." Okay. What if we replace all of that empty space with something that we'll call "FIBER". Only instead of running all of the fiber to the satellite and back we could just run it over land. Barring any service interruptions by 30 story lizards breathing fire all over the data center this might just work!
  • This means that new series are available to fansubbers even sooner than previously ! Yarrrr !

  • by Art Pollard (632734) on Saturday February 23 2008, @06:40PM (#22530296)

    With all this reliance on satellite technology for GPS, communications, and weather prediction what happens when (not if) the sun hits a more active solar cycle eliminating all of these satellites in one fell swoop? We have become terribly dependent on satellite technology (that I agree is cool). However, there have been solar storms that would knock out all of our satellites in recent memory -- only we did not have any satellites up yet. Now the satellites are up and the next large solar storm is just lurking out there getting ready to strike.

    As usual, beware any significant reliance on any one technology.

  • by jmf (30234) on Saturday February 23 2008, @07:00PM (#22530444) Homepage
    The name of the satellite has been mistranslated: 'kizuna' () means bonds [msmobiles.com] (as in 'family bonds') and not 'winds', which makes a lot more sense given the satellite's function.

    • Re: (Score:3, Informative)

      I had thought the same thing, but that's not correct. If you look at the JAXA page on the nickname [www.jaxa.jp], Kizuna is the nickname and the official name is WINDS (spelled all uppercase) which is an acronym. It's very confusing though.
  • by DrBuzzo (913503) on Saturday February 23 2008, @10:12PM (#22531850) Homepage
    You can get a substansial amount of bandwidth from satellites using spot beams and multiple channels on the Ku and now Ka bands. But that's not really all there is to it. A geostationary satellite (which is pretty much what you're stuck with unless you want tracking antennas and multiple satellite networks) is pretty high up, more than 30,000 miles. For the signal to get up to the satellite and back down to the earthstation then back the other way takes some time just because of the speed of light. The two-way pingback time can be as high as 400 miliseconds.

    For web surfing and stuff this can be somewhat compensated for by prefetching and caching to make it less noticable, but for VOIP or other realtime use the lag can be very very annoying to say the least and it reduces QOS and transfers in general. Especially for any kind of distributed networking, torrents or multiple file download, the lag can add up and result in service issues. A single lost packet can take a half second to report to the server and have resent as can a next-file request.

    These are not insurmountable and you can still get okay internet use for general purpose stuff, but it is a big disadvantage which there is really nothing you can do anything about. The speed of light is fixed.


    Satellites really shine in certain roles, especially broadcasting where you have point-to-multipoint distribution. They work very very well with one-way content or stuff that does not require dynamic two-way data exchanging. They also do pretty well for remote reporting. Satellites are not optimal for two way internet traffic. That's just the nature of the beast.
    • by Adambomb (118938) on Saturday February 23 2008, @04:27PM (#22529368) Journal
      Exactly, While this could be useful for bulk mobile file transfers, this definitely wont be used for anything real time.

      I believe geosync orbit has a MINIMUM lightspeed latency of 119.4ms.

      Not a fun starting point BEFORE collisions and noise.
      • Re: (Score:3, Informative)

        Ehm, you also have to get back down, so that is 240ms minimum...
        • wow, good call.

          and another tally on the mind-is-dying meter.

          =)
          • by dgatwood (11270) on Saturday February 23 2008, @05:29PM (#22529800) Journal

            You have to multiply times four to get a useful figure. Latency is normally measured round trip. Hop up, hop down, return hop up, return hop down. Latency to geostationary orbit is half a second.

            However, 175 miles up is NOT geostationary. Geostationary is 35,786 km up, give or take. The orbit is geosynchronous. That just means the orbital period is the same as the earth's rotation, so it returns to the same spot at the same time every day. It will NOT stay in the same place, however. They'll have to have several of these things in a similar orbit flying over periodically like we do for GPS satellites. It also means the round trip latency is about 3.76 msec (just less than a millisecond per hop), a heck of a lot shorter than half a second.

            • Re:Now featuring... (Score:5, Interesting)

              by absoluteflatness (913952) <absoluteflatness@nOSpAM.gmail.com> on Saturday February 23 2008, @06:03PM (#22530018)
              Geosynchronous and geostationary orbits are obtained at the same radius from the Earth, about, as you say, 35,786 km above sea level. The defining factor that differentiates between geosynchronous and geostationary is the plane of the Earth the orbit is positioned over. A geosynchronous orbit that is directly aligned with the Equator is "geostationary" since it will always stay above the same position on the Earth. Plain "geosynchronous" orbits are simply aligned differently.

              From the JAXA site about Kizuna:

              "Scheduled orbit: Geostationary orbit at 143 degrees East longitude and at an altitude of about 36,000 km"

              It is, even though the summarizer slipped up a bit (technically the term is correct, but somewhat misleading), destined for geostationary orbit.
              • Re: (Score:3, Interesting)

                lpq, you're smarter than the BBC. ;)

                First, the speed of light is slowed down by fibre optic cable, just as light is slowed travelling through any medium. Roughly light in optical fibre travels 2/3 the speed of light in a vacuum.

                So, to compute minimum latency, take the length of fibre, divide by the speed of light, divide by 2/3, and then double it, as the data must go there and back.

                Thus, if we had a 1300 km cable: 1300km / 299,792.458 km/s / (2/3) * 2 = 0.013009 s = 13 ms round-trip-time.

                So, for our 1300
        • And assuming you're talking about a round trip, which "latency" generally does, that doubles again to 480ms. Which is about the typical minimum ping time for your run-of-the-mill satellite internet these days. Switching delays are fairly negligible on that sort of a timescale, even with huge numbers of connections provided you're using fairly modern technology. Look at the cellphone networks: this is a solved problem.
        • Ehm, you also have to get back down, so that is 240ms minimum...
          Still couldn't this provide some measure of redundancy in cases where the hard lines are damaged or taking down for whatever reason, like we recently saw in the middle east?
        • Australia has 250ms lag to the US minimum.
          I'd sure prefer their speeds over 512k though. ;)
      • Bulk mobile one way, say, to an offsite backup server farm. That 155 Mbps looks awful tasty on the download link. Should be able to move lottsa pr0n...

        The 5 Mbps uplink is kinda weak, though. Forget about bittorrent...

    • Super Latency

      I dunno. If I had the choice between no internet or satellite, I wouldn't complain about the latency.

      Of course it might suck if it rains a lot, but I suppose it is far better than not having any internet. Secondly, if you are a SE Asia islander or boat traveler you might not even have dial up seeing there is no fiber to your location. You might have a LAN line, but it might be incompatible or really slow seeing regular modems don't work well with satellite phones.
      • I can get dialup or satellite. I can't play a game over the dialup anyway. I usually get 26.4kbps. Good times. I'm planning to get Hughesnet, which is the ONLY one that will give you a decent transfer allotment. Or so they say. Everyone else cuts you off pretty low.
        • Re: (Score:3, Interesting)

          I don't know what is worse..

          The fact that you posted this racist crap in the first place or the fact that you posted anon so you could mod down anyone that responded to you.
                • Unless you're pushing the button from your newly colonized extraterrestrial planet with a nice self-sustaining biosphere of its own =)
    • by Dunbal (464142) on Saturday February 23 2008, @06:25PM (#22530178)
      With radio waves traveling at the speed of light what difference is 175 miles going to make?

      175 miles? Try more like 22,230 miles. That's pretty much the only place you can put it unless you want your internet connection to only work 3 minutes out of every 90 minutes...

      The reasons are simple physics. Gravity causes everything to want to fall towards the center of the Earth. Satellites manage to stay in orbit because they are constantly "falling" ahead of the Earth. That's why things in "low earth orbit" are referred to as being "in freefall" and not REALLY in zero gravity. Gravity is still there, only the velocity of the satellite is so high that all gravity manages to do is curve the trajectory of the satellite, not cause it to lose height. This means your satellite is going to be moving VERY fast with respect to the ground.

      It's only at 22,230 miles out where the circle is so big that your satellite now appears fixed with respect to the ground. It's still moving. It's still "free-falling". But it appears to be hovering over a fixed spot over the equator - very useful for communication satellites since now you know where to aim your antenna and you don't have to bother moving it.