Forgot your password?
typodupeerror

Wired and Wireless At the Same High Speed 110

Posted by Zonk
from the zoom dept.
Roland Piquepaille writes "The next generation of optical networks needed to satisfy our appetite for bandwidth is currently under development. And researchers from Georgia Tech have built a new architecture which delivers super-broadband wired and wireless service simultaneously. This hybrid system 'could allow dual wired/wireless transmission up to 100 times faster than current networks.' In fact, this optical-wireless network can carry as many as 32 different channels, each providing 2.5 gigabit-per-second service to your home or your office. And companies such as NEC and BellSouth are already working on such hybrid optical-wireless communications networks."
This discussion has been archived. No new comments can be posted.

Wired and Wireless At the Same High Speed

Comments Filter:
  • WiMAX? (Score:4, Interesting)

    by Landak (798221) <Landak@gmail.com> on Saturday March 18, 2006 @07:50PM (#14949970) Homepage
    Is this going to be the successor to Intel's somewhat vapourware "WiMAX [wikipedia.org]" project - or is it this in all but name?
    • Re:WiMAX? (Score:1, Interesting)

      by Anonymous Coward
      WiMAX isn't vapourware. It's just been officially approved and commercial products are starting to hit the shelves
    • by j1m+5n0w (749199) on Saturday March 18, 2006 @10:01PM (#14950436) Homepage Journal
      Wimax is for city-sized networks. I wouldn't expect this new technology to work well over long distances or in bad weather; one of the articles indicated they were using milimeter-wavelength frequencies, which puts it somewhere around 100Ghz, which is stopped by water vapor. Wimax uses much lower frequencies (with correspondingly lower data throughput) that can (to some limited extend) go around corners and penetrate fog and rain.
      • The data capacity of a wireless system actually isn't dependent on the transmission frequency (the 100GHz frequency you specced for the first technology). It's dependent on the signal to noise ratio (which can depend on where in the frequency domain you're operating), and the available bandwidth (range of frequency on which you're transmitting data). A signal with 6MHz of bandwidth (TV channel sized) can carry the same data if it's mixed to 100KHz as at 100GHz (given the same SNR).
      • If bits are encoded on a wave then wouldn't more waves (higher frequency) give the opportunity to encode more bits and therefore increase throughout? Assuming the same SNR?
  • by CustomDesigned (250089) on Saturday March 18, 2006 @07:56PM (#14950004) Homepage Journal
    Because the capacity of optical fiber is so high, this optical-wireless network could use wavelength division multiplexing (WDM) to carry as many as 32 different channels, each providing 2.5 gigabit-per-second service. That would allow users within buildings to subscribe to services from many different providers, each with their own content.

    At first I was confused, because the article seemed to be talking about internet access. But then I noticed that Bell South was one of the sponsors. So, welcome to the future of the internet as envisioned by Bell South.

    • So, welcome to the future of the internet as envisioned by Bell South.

      You mean AT&T. Or whoever buys them out.
      • Doesn't matter what you call them. All the baby bell's are buying each other back up to reform the giant monopoly. And then they'll be broken up again, and the cycle repeats. Looks like the cable companies are getting in on the act as well. But I digress...

        I think (hope) what was actually intended by the abovementioned bold text was that your router would spit out signal on 32 different channels (and each computer could have its own, until you've got way too many machines connected to your home networ

    • Exactly. I was hoping I'd see the name of a company that might be able to offer some competition. So far, same 'ol same 'ol.
    • And Bell South was just bought by AT&T...

      (Cue: Empire Strikes Back Soundtrack...)
    • Because the capacity of optical fiber is so high, this optical-wireless network

      Well, except the new free space optics solution -- which isn't new per se, companies have had niche products for years -- isn't constrained in an optical fiber which would prevent interference. Instead it uses free space, and this immediately limits your bandwidth as you ultimately have to share it with people nearby you or nearby the other end of the air interface.
    • Suprise! Private industry actually funds research at public universities..! Some would say its downright beneficial to both parties. mic check. mic check, 1. 2. Like all good fellows I drink my whiskey clear I'm a ramblin gamblin helluvan engineer!
  • by nbert (785663)
    So if this new technology is going to be 100 times faster on both mediums why are they planning to use wires at all?
    Maybe it's because the wireless solution will suck so many frequency bands that it can't unleash it's full power unless you are living in a really remote area where other APs are quite unlikely.
    • Re:catch? (Score:5, Interesting)

      by TubeSteak (669689) on Saturday March 18, 2006 @08:21PM (#14950105) Journal
      From TFA:
      amplified for short-range wireless transmission at frequencies of 40 to 60 gigahertz.
      Anyone want to comment on how useful a 40~60 GHz signal is outside of a relatively controlled environment?
      • Re:catch? (Score:5, Informative)

        by erick99 (743982) <homerun@gmail.com> on Saturday March 18, 2006 @08:28PM (#14950137)
        Pure line-of-sight and signals at those frequencies are absorbed by all sorts of things including tree leaves (of all things). You need a really straight shot from transmitter-to-receiver. You also cannot run a great deal of power at those frequencies which can affect range. We play around with gigahertz range transceiving in ham radio and there are a lot of variables to take into consideration. I imagine they have so far tested it mostly in fairly ideal conditions(?). Erick KE3PB
        • Re:catch? (Score:2, Insightful)

          by cciRRus (889392)

          Pure line-of-sight and signals at those frequencies are absorbed by all sorts of things including tree leaves (of all things).

          Tree leaves are not really an issue since there is "pure" LOS, no blockages. However, the issue with such high frequency transmissions is the EM wave energy being absorbed by the water molecules present in the atmosphere. This presents a high level of signal attenuation which greatly reduces the range of the wireless tranmitter.

          Imagine you get the full data rate before it rains,

        • Sounds to me like the 5 mile walkie talkies that can function up to half a mile in ideal conditions. What ever happened to using more than a few percent ( 10% ) of the optical fiber laid. At least in a sheathed situation, the fiber is guarenteed to carry a significant amount of data rapidly. I'm also not sure how effective working on improving bandwidth is when the current tech is as faulty as it is. When I was in texas, I was across the street from my broad band provider, and had regular net outages s
      • If you go all the way up to 60 you hit a band which is absorbed by oxygen molecules. Signals don't go very far.

        Before then you're in a range that the military has used, at least experimentally, to image runways when landing in fog.

        Think short ranges (1 km for sure), shorter in humid environments, and a relatively benign interference environment since there are so few natural sources in that range and it's so easy to make a small highly directional antenna.

        People have only been holding off on deployments bec
      • So we are going from a wave length of 1500 nM to about 6 mM, a large expantion. The wave guide would be 1/8 by 1/4 inches , or antenae 1/8 high. Have a good cleaning crew in a public space. It is just that I remember my father's tales of sailors chipping paint off the wave guides in WWII. It may survive dust motes but dust bunnies will kill it.
    • nbert, if I seem rude, please excuse me. I simply assume you don't know much about the various medias. The reason for both the wireless and wired media is simple. Wireless is amazing in terms of how it works through sound waves, but try sending a signal through a cement wall located between you and the access point (the antennas that pick up the wireless signal). Worse yet, TWO cement walls! Plus the sound waves will eventually become weaker as it goes further and further. With fiber, however, you can use
      • um... those would be RADIO waves... not sound waves. Which sort of follow similar rules (being waves and all), but are definitely different...
        • wow, thanks for catching my mistake. Seriously, no sarcasm involved. Sound waves is the compression of air via vibrations; radio waves are caused through electromagnetic radiation.
        • um... those would be RADIO waves... not sound waves. Which sort of follow similar rules (being waves and all), but are definitely different...


          I don't know, I think a wireless LAN based on sound waves might be kind of neat. At the very least it could be used to drive the local users (and/or their dogs) crazy...

          • been there, done that. 75 baud data from one side of a room to the other through purely acoustic media. You could probably go a _little_ faster, but ultrasound suffers from echoes and multipath interference so the bandwidth is seriously limited.

  • layer? (Score:3, Insightful)

    by convolvatron (176505) on Saturday March 18, 2006 @07:59PM (#14950010)
    isn't this providing media interoperability at the wrong layer?

    the framing and termination guts of the wireless transceiver aren't all that expensive. there are already perfectly good layer 2 and 3 approaches to the problem of distributing the same content over wireless and wired networks'
  • by Aphrika (756248) on Saturday March 18, 2006 @07:59PM (#14950014)
    Won't more critical technologies limit how fast we can transmit data, such as switch fabrics?

    To effectively use incredibly fast end-user technologies, some absolutely incredible switches and routers would need to be designed, otherwise all this is for nothing. I mean 2.5 Gb per port on a 24-port switch would require a 60 Gb backplane - way higher than anything available today.

    And as someone who managed a medium-ish sized network (250+), we currently find that setting a lot of peripheral users to 10-full gives much better performance than setting them to 100-full, simply because our switching fabric - coupled with the number of users - can handle this a lot better.

    So although this is possible, wouldn't it be more suited to backbones, rather than having a client-heavy network?
    • True, but... (Score:5, Insightful)

      by John.P.Jones (601028) on Saturday March 18, 2006 @08:24PM (#14950121)
      Given there will be bottlenecks but I want them far away from me. I am sick of the last mile to my house being the bottleneck, move it somewhere else for a while, somewhere where it can be more easily updated.

      Now, once we have a wire to my house capable of some outrageous speed go ahead and restrict it to match your network speed as long as that excess capacity is kept in reserve for future improvements. This seems to me a more sensible way of engineering the network, the most expensive upgrades (last mile) should be done right once and let the rest of the network catch up after many incremental updates.
      • Ah, this doesn't make a whole lot of sense...

        If your house isn't the bottleneck, than as soon as the "bottleneck" is removed by upgrades, then what eventually becomes your house? [Hint: it will be last to receive the upgrades] ...

        Which is cheaper? 1) Install fiber to 100 existing homes [requiring digging,etc]? 2) Upgrading the line that runs between a neighborhood and another station, ONCE?

        What if it isn't even fiber? What if it just means replacing a device near each customer versus replacing equipme
      • I am not an engineer, but... For new areas, the price of laying in copper or optical cable is about the same (about 1500 dollars). In fiber optics, the speed could be 1Gb, in copper, about 24Mbit (currently?). Now, why would a phone company still make all the new areas with copper instead of fiber optics? Equipment might be more expensive for optics but still...
    • by Anonymous Coward
      "I mean 2.5 Gb per port on a 24-port switch would require a 60 Gb backplane - way higher than anything available today."

      Really?

      http://www.cisco.com/en/US/products/ps6421/prod_bu lletin0900aecd8036889f.html [cisco.com]

      "Full bisectional bandwidth for all ports, providing 2.8 Tbps (Cisco SFS 7012) and 5.4 Tbps (Cisco SFS 7024) of bandwidth"
    • There are core routers that can do over 300Gbps. Here is an example: http://www.juniper.net/products/tseries/ [juniper.net]
    • You must not work in the storage world... We have plenty of switches with a backplane that can handle 500 Gb/s... What you speak of is nothin' for a director class switch.
    • Actually network equipment vendors are getting pretty close to 60Gbps backplane speeds for individual line-cards.

      Both Cisco, with it's76xx series [cisco.com], and Force10, with it's Exxx series [force10networks.com], currently offer line-cards with 40Gbps switching capacity.

      Regards.
    • Won't more critical technologies limit how fast we can transmit data, such as switch fabrics? To effectively use incredibly fast end-user technologies, some absolutely incredible switches and routers would need to be designed, otherwise all this is for nothing. I mean 2.5 Gb per port on a 24-port switch would require a 60 Gb backplane - way higher than anything available today.

      No offense, but what the hell are you talking about? The low-end managed switch fabrics have as much bandwidth as you are discus

    • Switch fabric forwarding capacity is usually measured in pps not bps. They are cut-through and the size of the packet in general shouldn't matter. This is true for copy-once type switches (i.e. Nortel 8600). where the packet destination is determined and copied to the output queue only once. Some switches don't do this and packet size comes into play much more. Our switches can do 7 billion pps on the backplane, which means they top out at 10 GBps (GigaBytes ps). Compare that to a legacy router than c
    • You can get an InfiniBand switch for 24 10 gigibit ports at full speed from SilverStorm, Cisco, Voltaire, HP, IBM, etc. All are based on a mellanox ASIC and probably under $1000/port. I found an HP box for 6500 on froogle just now. I am sure there are boxes at this level for eternet switching and ip routing. Check out this monster [foundrynet.com] from Foundry. From here [foundrynet.com], it has "3.84 Terabits per second of backplane switching capacity".
      -Ack
  • But... (Score:5, Funny)

    by Anonymous Coward on Saturday March 18, 2006 @08:07PM (#14950044)
    Can it run a Beowulf Cluster of Soviet Russian... ah fuck it
  • my office? (Score:5, Funny)

    by jjeffries (17675) on Saturday March 18, 2006 @08:15PM (#14950076)
    Did anyone else see the picture in that first link and think "Hey, what's that dude doing behind my desk???"

    • Nah, you can't see the lower strata of cable running across mine for the mountain of paper which has formed over it. I expect to create fossil fuels over time by using the heat from inefficient cabling and the pressure from a 274 page status document on my old data flowcharts.
    • Saw pile of switches and I thought that. But there was no pile o unix workstations... so it cant be my desk.

      Also looks way too organized.
    • Nah, I was like, "how the hell does that guy get any work done without a damn monitor?!? And why the HELL does he need an umbrella in the office?"
  • by sglider (648795) on Saturday March 18, 2006 @08:16PM (#14950081) Homepage Journal
    I could care less how fast the speed is. 192kpbs is currently how fast the fastest multiplayer game operates. I care about latency. Fix that problem and we'll talk.
    • by Anonymous Coward
      The speed of light isn't good enough for you?
      • Not if it's bouncing off a satellite in geosynchronous orbit!
        • Not if it's bouncing off a satellite in geosynchronous orbit!

          Which no internet links the average consumer will use does, it adds a 280ms hop. The main overcomable problem is the time taken to convert light to electricity, switch, and convert back to light. Switches that use mirrors and can switch light directly will save time.
    • The only way I can think of is to use an array of trapped Entangled photons, that can be switched at either end. That would remove the largest bottleneck in latency between major sites, e.g. ISP to data centre where the game servers are located. And in theory playing on the other side of the world could be as fast as playing within the same country.

      Going off on a tangent, does anyone know if there's a way to detect these photons compared to non-entangled ones? I only ask because surely for SETI, it would
      • Ummm... as far as I know, even with entangled photons, you cannot transfer information faster than the speed of light. Now, that method may speed up the transfer by a very small amount, but most lag is created by switches and routers, not the time it takes for a signal to go through the cable (at least, as long as you are not going across an ocean or bouncing off a satellite).
    • Latency in non-overloaded devices is a function of the speed of light. Perhaps you can tell us how to transmit information faster than light.

      http://www.rfc-editor.org/rfc/rfc1925.txt [rfc-editor.org] is an important read.
  • If this were to actually work and be rolled out(said with a GIANT grain of salt), what traveler in an airport with their laptop has that kind of HD space? I'm assuming if you're going to have a connection that fast, it would be to watch HD movies and things of that nature. Laptop HDs aren't cheap, and is there truly a large scale need for something like that?
    • What's this "hard drive" you're speaking of, when you have access to that kind of bandwidth? (Oh, and I'm clearly in the "I want to own my own data" camp, just not in the "streaming is impossible" camp...)
  • WARNING! (Score:3, Informative)

    by Toasty16 (586358) on Saturday March 18, 2006 @08:23PM (#14950119) Homepage
    This article has been submitted by Roland Piquepaille [slashdot.org], proceed to the linked articles with extreme caution!
    • This article has been submitted by Roland Piquepaille, proceed to the linked articles with extreme caution!

      So?

      a) Primidi.com is only linked through rel=nofollow in the article summary. This means that Google doesn't increase his pagerank.

      b) There are no links to his personal website in the submission. The linked articles are not on his website; therefore, all your criticism about Blogads and whatever else is moot.

      c) Everyone links their name to their personal website when submitting stories. Just because R
      • Re:WARNING! (Score:3, Informative)

        by Anonymous Coward
        One of the links is to his ZDNET blog. He is avoiding criticism by "existing" at primidi but blogging at ZDNET, however nothing else has changed.
      • Damn dude, I was just trying to get a "+5 Funny". Sheesh... ;-)
  • Wonald (Score:4, Informative)

    by Anonymous Coward on Saturday March 18, 2006 @08:55PM (#14950234)
    One link to Wonald's ZDnet blog, 6 links from there to his link farm, with up to 10 links per page to other Wonald blogs. Remind me again how much Wonald pays Slashdot for his slashverts?
  • weather? (Score:5, Interesting)

    by j1m+5n0w (749199) on Saturday March 18, 2006 @09:14PM (#14950292) Homepage Journal
    This sounds like free space optics [wikipedia.org], which in bad weather is only reliable over short distances. This could very well be interesting technology, but my enthusiasm will remain subdued until I hear how well it performs through, say, several hundred meters of thick fog.
    • The article refers to millimeter waves (40-60GHz), this is true line of sight stuff.
      Several hundred meters if an order of magnitude more range than I would expect
      it to be disigned for.

      This short range is an advantage, and a disadvantage. The network would have to
      be 'cellular' in nature. There would need to be several base stations per area,
      even someone standing between you and the base station would block the signal.

      On the other hand the re-use distance for a given channel would be fairly small,
      the band is
    • until I hear how well it performs through, say, several hundred meters of thick fog.

      Probably like a sniper with a $4000 telescopic sight.
      (Just think about it)

    • >through, say, several hundred meters of thick fog.

      Millimeter wave frequencies penetrate fog splendidly but are blocked by rain. Since that's sorta the opposite of what visible light and infrared do, a line of sight system could get good reliability by having both optical and mm-wave components and some failover logic.
    • but my enthusiasm will remain subdued until I hear how well it performs through, say, several hundred meters of thick fog.

      You're from the UK, aren't you...

      -Chris

    • This sounds like free space optics

      Are you just meaning it's faster than what we currently use? otherwise it doesn't sound much like FSO. It seems to me that their wireless component is nothing special. Higher than current frequencies, not as high as FSO, but similar to what many others are working on.

      The interesting part is the hybrid feature. Sounds like they are creating a single physical layer where the signal travels on both wireless media (the air) and wired media (optic cable) using an all optical c

  • by zappepcs (820751) on Saturday March 18, 2006 @09:18PM (#14950306) Journal
    Seems to me that one of the sponsors of this tech is Bell, and aren't they the ones that want to charge us for guaranteed latency, or lack there of? With all that bandwidth, that makes charging extra for low latency a case of banditry, doesn't it? Perhaps that is what Bell is all about anyway. On the other hand, I thought part of the reason for a tiered Internet service was to pay for all the infrastructure that is currently built? Now they are building 100x infrastructure with the money they are already overcharging from users, and only to overcharge them for content they don't want or need in the future?

    Sure, I'm not Mr. Optimistic here, but just who the hell is paying for this infrastructure? Already I only want 35% of the content I have to pay for, and none of what I pay for has the latency that I would like to have. The money vs. service issue is all out of whack here. I don't care if its wireless or wired personally, if they could just get the service right in the first place, it would be nice.

    Bundled cable, ISP, and VoIP... this is starting to sound like the beginnings of Cable Operators part two. I just know that they need all the bandwidth to support the DRM content that nobody wants to pay for, never mind watch. All I need is DRM'd reruns of "I love Lucy" on my telephone bill to make the world a perfect place again.

    There is simply way too much HYPE in the technology sector these days. God forbid any of them think of providing good service before figuring out how to sell me 2 terabits of bandwidth to watch reruns with.

    I'm not feeling very enthused about ISPs and content industries right now...

  • Someone ping me when...

    1: Realtec has a $19.99 add on card with technology
    or
    2: It automatically comes with my next motherboard purchase

    Untill that time, I'm plenty fine with the 6mbits/sec that I'm currently getting from the my internet provider.

    BBH
  • Woooooohoo!!! (Score:1, Interesting)

    by Anonymous Coward
    It will take another 40 years before that high speed network will be available in the good ol USA

    Meanwhile, Japan is getting 100Mbits/sec at $30/month
  • So by 2.5Gbps, they mean down right? And upload is obviously limited to 100mbit... er... what... the internet doesn't really work that way after all?
  • IBM has promised us something even greater. They are calling it "Token-Ring".

    Start buying your tokens on EBay now. You can never have too many!
  • Someone to discover a way to take advantage of duality of light and start transmitting internets wirelessly through light waves...
  • It's too bad that the paper isn't available online, because it would be interesting to read the source documents. Descriptions of technical papers intended for general audiences often lose quite a bit in translation. As an example, it's hard to tell what the article means by "100 times faster than existing networks," as an earlier poster pointed out (I'd guess the comparison is to gigabit ethernet, as 2.5 x 32 = 80, which is sort of like 100). Researchers always know their subject better than reviewers,
  • I'm having a hard time seeing the point of it. Do you want to be able to unplug your computer from a wired network, and still have it communicate? Or, do you want to be able to plug your laptop into a wired network and have no interruption of data communications? While this technology would seem to accomplish a seemless switch between the two, I don't know that such a goal is something that the marketplace of ideas is looking for. It is intriguing, but I'm not sure that its something people are looking
  • I didn't care about the story, but it was submitted by Roland Piquepaille and I came in here to read the fighting over him, all his submissions, etc. And there's almost nothing here. Where is all the anger today? Slashdot, you disappoint me...

Stupidity, like virtue, is its own reward.

Working...