Forgot your password?
typodupeerror
Wireless Networking Communications Hardware

Scientists Test World's Fastest Wireless Network 77

Posted by Soulskill
from the what-can't-lasers-do dept.
MojoKid writes "Scientists in Pisa, Italy claim to have set a new world record for the fastest wireless data transmission. They report that they were able to achieve throughput speeds above 1.2 Terabits per second, which they say beats the previous wireless data transmission speed record of 160 Gigabits per second, achieved by Korean scientists. The technology that the Pisa scientists utilized actually shares a significant similarity with fiber optics. Unlike Wi-Fi or microwave communications, which use radio-based transmissions, the Pisa scientists used a technology called free-space optical communications. In free space optics, an energy beam is collimated and transmitted through space rather than being guided through an optical cable."
This discussion has been archived. No new comments can be posted.

Scientists Test World's Fastest Wireless Network

Comments Filter:
  • by Dr.Pete (1021137) on Saturday September 13, 2008 @10:14AM (#24990109) Journal
    While technically achieved without wires, the thing with WiFi is that you can used the omnidirectional nature of the transmission (along with refraction and diffraction of the signal) to access the signal anywhere within its range, often without line of sight. Naturally, free space optics requires an uninterrupted line of sight and significant alignment procedures. Now I'm not saying line of sight networking is useless (it was used to great effect after 9/11 and is great for places you need a temp. network but can't string a wire) but comparing it to radio WiFi is a bit apples/oranges.
    • by xZgf6xHx2uhoAj9D (1160707) on Saturday September 13, 2008 @10:39AM (#24990323)

      They weren't comparing it to WiFi specifically. They were comparing it to wireless as a whole.

      It is "really wireless". It doesn't use wires.

      • By that token Fibre Optics are "wireless" due to using an optical cable, rather than a metal wire.

        The article heading is slightly misleading.

        Either way, in practice, this isn't the best for communication as if LOS is broken, so is the data transmission. Simple atmospheric effects could lead to some pretty crap throughput.

    • Re: (Score:3, Funny)

      by Adambomb (118938)

      Sounds like a good replacement in terms of energy use for microwave bridging and other methods that require LoS already.

      Of course, packetloss due to geese...

    • by gerf (532474)

      There are limits to this method, but it does have some possible implementations. Maybe we'll use directional optical links to and from satellites that need huge amounts of data transferred (Hubble/spy/moon colony).

      Also, this is a cool breakthrough for the sake of science and progress in general. Though I normally join in cynicism, I can actually see some benefits in this experiment.

    • My company uses a LOS wireless device for our main internet connection. It's cheaper than a fiber link and gives us decent bandwidth...

      we're on the 4th floor of an office building and happen to have an open window pointing in the right direction though... so YMMV.

  • by Anonymous Coward on Saturday September 13, 2008 @10:14AM (#24990113)

    I always knew those scientists in Pisa leaned toward having faster networks. They've always towered over the rest of the world, technologically.

  • Really (Score:2, Interesting)

    by cefek (148764)

    there should be a difference between point-2-point speed record, and point-2-multipoint, which concerns most of us wifi-users. P2P connections are used mainly in business, and they can have backup links should, let's say' fog happen that will disperse all the transmission (reduced wisibility).

    • Re: (Score:2, Interesting)

      by Bios_Hakr (68586)

      Couldn't a rotating prism be used to turn this into a miltipoint link? Sure, you'd lose some bandwidth and you'd still need LOS. But you could easily queue and time packet transmission in time with the rotation. Kinda like old airplanes shooting bullets through the blades.

      • by cefek (148764)

        I'm not really certain, but if you multiplex the signal, don't you sort of lose control over it? I mean, if it's easier to spot, it's easier to sniff. This would, however, solve the problem of bird flock blocking the 'net access for the whole district. So it really depends.

  • Free space optics has been around for a while, and cost has been more of a stumbling block than speed. I'd much rather have a 10 mb/s rig for $500, rather than the 1000 mb/s rigs companies are selling today for $50,000.

    I've been interested in the Ronja project for a while, but it's very labor intensive to build and deploy. Somebody ought to commercialize it.

    • Ronja is pretty much on topic. http://ronja.twibright.com/ [twibright.com]

      From their FAQ:

      Material for one Ronja 10M Metropolis device costs 2000CZK and building the device takes 70 hours.

      2000CZK; about $120. 70 hours though is even at minimum wage ($6.55) $500 or so. You need two of them, so really a minimum of $1,240 in costs. Then add markup, even here the devices are going to cost $2,500 just for the hardware to set up a link, add consultancy, site surveys and you're into the same ballpark as the existing commercial FSO providers.

      What is off topic though, the fact that same bunch of people seem to have largely solve

      • by TheLink (130905)
        "70 hours though is even at minimum wage ($6.55) $500 or so"

        That's why stuff is made in China :).

        And that's why stuff like my radio controlled heli costs less than USD20.
        • by Colin Smith (2679)

          Sure, but to do that you have to be able to sell millions of the things. For FSO, the market just isn't that big. How many people need to network buildings which are = 1.2km apart? The fact that there aren't FSO devices sitting on best Buy shelves beside the wireless routers speaks volumes.

           

  • by AB3A (192265) on Saturday September 13, 2008 @10:20AM (#24990163) Homepage Journal

    ...the only way you'll get it is to move up in the electromagnetic spectrum. It had to be a laser based communications system.

    The alternative is to smear this crap all over the electromagnetic spectrum. And at this data rate, if you really expect throughput, you can't rely on spread spectrum to save you.

  • by yttrstein (891553) on Saturday September 13, 2008 @10:22AM (#24990173) Homepage
    I was about to say the same thing about wireless being omnidirectional and "free space" optics being single point source and directional.

    It occurs to me however that this is not new technology. Back in Chicago in the mid 90s, I did some interesting work at a very well funded mom-and-pop ISP that was playing with some "line of sight" (RF and optical) T-1 equivalents. The "free space optic" portion of the circuit died completely every time it rained, so it wasn't too terribly useful for anything outdoors, like shooting a high speed line across town by aiming a couple of transceivers out some open windows.
  • CMCps (Score:5, Funny)

    by xigxag (167441) on Saturday September 13, 2008 @10:34AM (#24990277)

    Another way to look at that is 6/10ths of a Comcast Monthly Cap per second.

  • It would be interesting if this technology could be used in a computer to make super fast buses.
  • by Announcer (816755) on Saturday September 13, 2008 @11:04AM (#24990493) Homepage

    They are all around us... laptops, PDA's, etc. Even your TV remote! Much lower bitrates, obviously. Essentially, these guys used a laser beam instead of a simple LED.

    Kicking it up to just above a Terabit per second is an impressive feat of technology. Unfortrunately, the range and throughput of such a system would be limited by various environmental factors- dust, smoke, and water vapor. Not to mention the power of the sun, outdoors. You also have numerous other sources of light pollution getting in the way. Using it within existing tunnels might be one way of avoiding stringing fiber through them... but it wouldn't take much to disrupt the beam. A spider's web in just the wrong place, for example, where the spider or one of its victims blocks the beam.

    Still, it's pretty cool to get the data that fast through the air.

    • by Acapulco (1289274)
      Well, you could always put some sort of "cover" to this laser beam, such that no spider can get in front of it.

      Right? :)
    • The devices I've seen (Ronja [twibright.com]) spread the beam out quite a bit, making it a fairly wide cylinder rather than a line. Atmospheric attenuation and sunlight are issues, but a spider web probably wouldn't block enough of the beam to make a difference. I would assume that other FSO equipment employs a similar principle.

  • It's not proper to call it wireless when there would be no wires in any case. It's glassless. And, as they imply, amplitude modulation of light is rather different than phase modulation of coherent RF.

  • by glindsey (73730) on Saturday September 13, 2008 @11:11AM (#24990549)

    they were able to achieve throughput speeds above 1.2 Terabits per second

    Unfortunately, what they didn't tell you was that all those bits were zeroes.

  • Whatever bandwidth you can get into an optical fiber is easy to transmit wirelessly point-to-point. A collimating lens on the end of the TX and RX fiber and you are done.

    To fill up the link with data, you need a 40Gb/s DWDM laser. You can easily add 128 different colors of these, and your link is up to 5 Tb/s.

    I needed a source of light like this to test a receiver/spectrum analyzer, and put one of these together (with the 10 lasers I could find muxed together), and did not bother to mention it.

  • Personally, I'm more interested in how they got data read from the storage medium at 1.2 terabits/s - unless it was random data?
    • I admit I planted this story just to get the cable/telcos to hurry up with that last mile stuff. You have to shock these people into thinking that wires will be irrelevant before they can get the money back on their fiber investment. Now start digging up the streets before I link to the article from all my ghost blogs, get it to the top of google news and bankrupt you with bits...uh...I guess I have to add...bee-atches!
    • by Khashishi (775369)
      The test runs were probably with procedurally generated data, something easy to confirm.
  • Never underestimate the bandwith of a wagon load of tapes dropped from the tower...

    Forgive me. (cowers)

  • They report that they were able to achieve throughput speeds above 1.2 Terabits per second, which they say beats the previous wireless data transmission speed record of 160 Gigabits per second by Korean scientists.

    And who could doubt them. I mean, who in their right mind could tell these scientists that 1.2 Terabits/second is not faster than 160 Gigabits/second?

    Sorry to be pedantic, but can't we try to be a little precise, sometimes?

  • Wow, that's amazing. They can deliver the internet faster and in such a capacity that I would be going over my monthly limit in uh.. 1/6 of a second! :)
  • What you need is to increase the power of the laser every time the signal drops. Say a goose flies into the line, you just need to increase the power of the laser (a lot) and you could burn right through the goose without losing any packets. With a powerful enough laser, nothing could stop the downloading of 1.2 TB of information at any given second of the day.
  • This story is intriguing but, as many have pointed out, has its obvious limitations. I thought slashdot had an article once about using plastics instead of glass for fibre optics. Why not invest more time and energy in this way? That would take the cost of fibre networking down significantly. I think, since a large cost savings could be realized, telcos would have a better chance of upgrading aging infrastructure and we might have broadband speeds that are more favorably compared in Japan.
  • It's amazing 1.2 Terabits per second. How did they deliver the data to the laser quick enough, even if they were just a bunch of zero's. Beats the pant off my Commodore 64 web-server. c64web.com
  • More important:

    10Mbit Wireless that doesn't require line of sight.
    10Mbit Wireless that has much more range. (eg, miles, not feet)

    Both of the above, 'consumer' friendly and Ethernet-compatible.

    Then work on a 54Mbit version.

  • despite the million dollars it takes to set up a connection like that, I want it.

Thufir's a Harkonnen now.

Working...