Please create an account to participate in the Slashdot moderation system

 


Forgot your password?
Close
typodupeerror
×
Wireless Networking Hardware Science

More Antennas, Faster Wireless 110

Posted by timothy from the aim-high dept.
rouge86 writes "The New Scientist has a story on how researchers broke the network speed record using a wireless network and multiple antennas. They plan to use the demonstration to show how powerful multiple antennas can be. Applications include power saving on mobile phones and reducing interference."
This discussion has been archived. No new comments can be posted.

More Antennas, Faster Wireless More

More Antennas, Faster Wireless

Comments Filter:
  • This is just another sign that wireless is replacing wired networks around the world. I'm guessing sometime in the near future wireless will outnumber wired networks. I think that everyone can be excited about this.
    • Its just so fantastic to be able to break into a network in a few moments and be on your way to a big payoff through corporate espionage with next to no work. You just have to love how technology is adopted in critical areas well before anyone has got off their ass to learn how to do it right.

    • Re:Wireless the wave of the future (Score:4, Interesting)

      by m50d ( 797211 ) on Sunday December 12, 2004 @03:58PM (#11067633) Homepage Journal
      Where this will become really interesting is when there are enough wireless networks that they all link up. Once that happens, there won't be any need for ISPs as we know them - just get your wireless box and join the big mesh that's out there. No connection fees, no censorship - then we will have a truly free internet. Transatlantic etc. links will be slower, but I'm sure that's a problem we'll overcome.
      • No connection fees and no censorship.

        No guarantee of quality of service.
        No tech support.

        Not to mention that SOME people will still have to have conventional ISP accounts, otherwise there will be no bandwidth for everyone else. These people will also likely need to have bigger and buffer hardware to handle being a data chokepoint.
      • There are afew basic problems with a wireless mesh network Routing, Bandwith becomes devided by the number of hops. And local bandwith devided by the number of users, and full coverage becomes limited.

        Now asuming you just wanted to talk to califona from new york and each hop got you 50 miles. That's... 2462 / 50 or 30 hops so your net bandwith would be 1/30th of the avalable bandwith asuming full coverage. Now chances are it's going to be closer to 100 hops aka not a strat line and not alwase 100 miles.

        • Now given to options pay nothing and get say 100th the bandwith or pay 50$ an month and get full bandwith most people are going to opt for spending cash

          NetZero: $10/mo for 48 kbps. Comcast: $46/mo for 3000 kbps. Trust me, people will put up with 1/60 the speed to save cash. And are you sure it'll be 1/100 the bandwidth, or just 100 times the latency?

          So unless the mesh networks start having servaces that the ISP wish to connect to your not going to see mesh networks making the internet free.

          Some un

          • Ok say MeshNet had a bandwith of 50kbps I think it would be around 1-30Kbps to the internet but that's a difrent issue.
            Now most people would pay for high speed of some sort 10,000 Kbps connections are out in some areas for 60$ a month so everyone that's still useing a modem by choice would stick with mesh net everyone else would have real internet acsess.

            A nation wide Meshnet would have horable bandwith but within a city it might not be that bad. Say you had 10,000 people on mesh net in New York at an
        • No, the bandwidth won't be limited, true it will be the lowest of those 100 hops, but wait, there are multiple paths from NY to california and the one you use will be the fastest one, so it will even out. Since there are no wireless standards with <11mbps bandwidth, it should be 11mbps all the way, as there is usually at least one 11mbps path between any two nodes. As for routing, nodes just need to know the general direction to route in. Intelligent routing algorithms are a problem, yes, but with sensib
          • You total bandwith would be the bandwith of each basestation.
            First off there is more than one person using the system so divide by the number of users.
            Second devide by the number of average hops because each message needs to use bandwith from more than one network.

            Now you have 1,000,000 users over 10,000,000 10Mbits/s stattions with an average hop count of 100.

            Total bandwith = 10,000,000 * 10 Mbit = 100,000,000 Mbits/s. Each users bandwith = 100,000,000 Mbits/s / 1000,000 / users / 100 hops. = 10KB/s
    • They said the same about phone networks when satalites were introduced. Now our world is connected by fibre. All of these measures, wireless, ADSL, cable modems are a stopgap for the fact that fibre is still too expensive to run into every house.

      If (and when) a breakthrough is made, either with an economic optical switch or the pricing of electronic processing equipment for optical fibre, wireless will once again take a back seat.

      d

  • 3G phones (Score:1, Insightful)

    by Almond Paste ( 838493 )
    Power saving is very welcome indeed. Today's 3G phones run out of battery in no-time.

  • Great engineering (Score:4, Informative)

    by Rosco P. Coltrane ( 209368 ) on Sunday December 12, 2004 @03:12PM (#11067386)
    For those who don't care about reading the entire article, the crux of it is:

    Recombining smaller signals in real time, however, requires considerable computing power. So the Siemens team developed new computer algorithms in order to send more data using existing hardware.

    In short: programmers managed to push existing hardware with a more efficient code. That's called hacking, albeit with a serious look, and I like that!
    • short and sweet, three cheers to the engineers... i guess good code on not-so-good hardware can do wonders too... like linux on a P3 machine compared to windows XP... what do you say?!
    • You can view the Siemens press release [siemens.com] (http://www.siemens.com/index.jsp?sdc_p=cfi1232554 lmno1232554ps5uz1& ) I'm not quite sure why this is news for slashdot. Maybe it's a quite news time. A couple of points:
      1. It's not existing hardware. It needs multiple receivers and multiple transmitters.
      2. They use 100 MHz bandwidth, so would expect to get a factor of 5 increase over the 54Mbps delivered by 802.11a simply through that.
      3. They use 4x3 MIMO which has been around for a while, see particularly the
    • I wonder if something like the Grid (similar to Seti@home and the like) could be implemented to contribute cycles to run the code. It could be one of the more interesting implementations, I would think.
    • Ok, now implement the more efficient code in hardware ;)

      (Seriously-- I don't like my devices passively using processor time)

  • duplicate post (Score:5, Informative)

    by Anonymous Coward on Sunday December 12, 2004 @03:19PM (#11067428)
    this is the same post, as the dec 08 slashdot post here: http://slashdot.org/article.pl?sid=04/12/08/022625 0&from=rss/ [slashdot.org]

  • OFDM Has Been Around for a While (Score:4, Interesting)

    by LuxuryYacht ( 229372 ) on Sunday December 12, 2004 @03:20PM (#11067435) Homepage
    OFDM has been around for a while OFDM History [ofdm-forum.com]

    It's nice to see more practical uses of it in wireless standards like WiFi IEEE 802.11a, 802.11g and in WiMax IEEE 802.16a.

    All this adds up to the death of the control by telco's in the last 100 yards of net connectivity. Go OFDM!!
    • OFDM has nothing to do with mulitple antenae. You can send OFDM signals across a channel with a single RX and TX antena. OFDM is simply a convinient way to turn a freq-selective fading channel into multiple, independent flat fading channels by interpreting the data as the IFFT of the TX signal. Spreading the data out over frequency can be used to get good diversity and thus avoid bad fades.

      Using multiple antenae is also a good way to get diversity in a way that complements OFDM (spatial vs. freqency).

  • hmm.. wavelets? (Score:3, Interesting)

    by Anonymous Coward on Sunday December 12, 2004 @03:23PM (#11067455)
    Since the improvement was mainly a hack on the processing.. I'd guess that they went from some sort of fourier transform to wavelets.. wavelets have linear computational complexity (awesome) and don't have the interference problems that older signal processing algorithms have.
  • Is anyone else making the comparison between "more antennas=better" and the mach 3 razor "more blades=better"?
  • ...now I have to buy a new phone again.

  • Interference (Score:2, Informative)

    by Barryke ( 772876 )
    Applications include power saving on mobile phones and reducing interference.
    Reducing interference.

    And increasing it for the neighbour, unless he also has multiple antennas.
    Wich gets us back to the start, only with even more interference...
    • Actually, what multiple antennas let you do is reduce multipath issues considerably when you combine them with DSP techniques. Two antennas doesn't increase interfereance- all it does is allow you to emit more RF power into the air effectively. Same goes for more than two.

      The main reason for not using multiple antennas has been cost and you really couldn't use them for recieveing in the past because of things like multipath playing merry hob with your reception. Nowadays, things are quite different.
  • Porcupines discovered this way before these researchers did.

  • So wait... (Score:2, Funny)

    by t_allardyce ( 48447 )
    Its not the size of your antenna, its the number you have??
  • Of course, this means more fodder for us wardrivers [wifimaps.com] -- more antennas mean more UINs to map!
  • I learnt a couple of years ago in Telecommunications class about the benefit of more than one antenna for wireless communications. Maybe the big deal is that products are starting to come out to the market. The technology however, is nothing new.

  • The principle is nothing new (Score:5, Funny)

    by Linker3000 ( 626634 ) on Sunday December 12, 2004 @05:54PM (#11068197) Journal
    I see nothing new in this principle - anyone knows that more aerials are better - in fact, if you increase the number of aerials so that their combined length (l) matches the exact distance between the sender and the receiver (ie: l = d), AND then you place each aerial in and end-to-end configuration so that electrons can flow in an unbroken path from transmitter to receiver then you have a very efficient data transmission medium - heck, the principle even works with optical fibre and photons too.
  • By contrast, the average wireless computer network can send only around 50 megabits (50 million bits) of data per second.

    I would think that a mean network would be the 802.11b network running 11MBits, not 50. Are there really that many 802.11g networks out there, pushing the average up to 50, or is this reporter just clueless because their office runs around 50? I used to travel quite a bit, and what networks I did run into at coffee shops and airports were 11 Mbit. Do others have a different view than

  • Wait... (Score:2, Funny)

    by Kurayamino-X ( 557754 )
    so now i need six more pringles cans for my end alone?
  • So we replace wires with wireless but double up on the amount of terminated wires (ie antenna's), cute. Aint you all glad somebody never put a RS232 on the back of a microwave oven, as I'm sure somebody would convert it into a crude basestation :)
  • So we get rid of all the wiring inside and terminated it on the outside a few doxen times instead and coat in plastic, and people get upset about mobile phone masts, boy are they in for a new shock soon when Joe Smith turns his house into a hedgehog array.
  • we all turn into Radioactive man. Or should I say... cancer man

  • This is already under consideration... (Score:3, Interesting)

    by Bored Huge Krill ( 687363 ) on Monday December 13, 2004 @12:37AM (#11070144)
    ...for standardization as 802.11n

    Proposals were submitted back in August for 802.11n, and all proposals still in the running use MIMO+OFDM (the technique described here). Hardware supporting various prototypes is already around in a usable form.

    It seems unlikely that 3x4 MIMO will be around in the first wave, due to cost constraints - 2x3 (2 tx, 3 rx) is the most likely initial configuration.
  • This reeks of channel-bonding to me. How (if it is) is this different?
    • it's different from channel bonding because the independent streams in a MIMO system use the same channel. It uses spatial diversity, not frequency diversity, to increase the channel capacity.
  • We can't disparage the telco's control too much. Though tyranical at times, there is a financial backbone that we don't want to break. We're pushing for a Marxian revolt on the telco's to go Robin Hood on internet bandwidth, but we're griping at an alarming rate about IT jobs going overseas and jobs being lost in the IT sector. We want the technology, but we don't want to pay for it...people want what you make at your job for free too.

    This is the same entitlement scenario that was looked at with MP3s an

Slashdot Top Deals

To do nothing is to be nothing.

Close