Stories
Slash Boxes
Comments
Search

News for nerds, stuff that matters

Slashdot Log In

Log In

Create Account  |  Retrieve Password

High-Speed Wireless LANs Move Forward

Posted by timothy on Mon Nov 27, 2000 03:11 AM
from the network-zip-11215-anyone? dept.
Technology
GrokSoup writes: "Neat article from the WSJ (free site) about European hackers using 802.11 technology to create limited-range, high-speed outdoor networks. As you might expect, people are messing with directed antennas to send signals up to a kilometer. While I've tried this to get from the house to the pool, the idea of banding together in open-source fashion had never really occurred to me. Nifty!" We've mentioned consume.net before, but this piece mentions some interesting possibilities, like how the same idea may result in an approved-by-the-Man wireless network in Sweden, and the golden hope that multiple connection methods will let us switch handily among several wireless protocols as the occasion merits.
+ -
story
This discussion has been archived.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.

High-Speed Wireless LANS Move Forward 50 Comments More /

 Full
 Abbreviated
 Hidden
More
Loading... please wait.

  • wireless wan in australia (Score:1)

    by Anonymous Coward
    Checkout Australia's effort on the same front.. www.air.net.au sp78
  • My dream is to be able to sit on the quad with a laptop and work on my online HW, yet be able to watch people play frisbee... Hopefully before i graduate...


    mov ax, 13h
    int 10h
  • YES! I want one too. I'll pay $50 for one!
  • why not modex? Or are you writing a 90-degree-rotated Wolfenstein clone (and who isn't)?
    ---
  • I'll be able to telnet into my toaster and hack into other people's refrigerators late at night to see what there is to eat.


  • I've heard about these kind of free wireless networks before, and it sounds pretty cool. One thing I don't understand is this: how do these wireless nets connect to the mainstream internet? More specifically, who pays for the link?

    I have a always-on connection, but my ISP would crucify me on the telephone pole outside my house if I started routing random wireless user's packets through their network.

    For an experimental setup, there are plenty of universities and the like willing to donate bandwidth. I don't see this scaling, though.
  • Wireless LANS are still not suited for real use. They can be eavesdropped. Apart from that, there are stories where the someone from across the street (with it's own wlan) logs in on the network in a building on the other side of the street (which was supposed to be high-security). They could see all domains etc etc.
    The case I remember was where some computer-magazine was testing wlans and they saw the entire domains of the hospital (or some other medical facility) on the other side of the street..
    I'm staying aways from these things...
    Plus I think there's enough waves going thru my head already..
  • my ISP would crucify me on the telephone pole outside my house

    That is right on par with high-school-dropout-sysadminning-for-pot-money comment I read earlier to-day. Thank you, good sir, for your most enter-taining witicisms. Right-O.
    ---

  • Anyone else near Clapham Junction fancy joining together to do this? I have a WLAN in my office.
  • That is the funniest sig I've seen!
  • I would but that would mean using Java.
    ---
  • Most WLANs support 40 and 128bit encryption. Eavesdropping is irrelevant. Unless you happen to be listening with your own quantum computer.

  • Finland has wavelan ISP's (Score:5)

    by nchip (28683) on Sunday November 26 2000, @10:40PM (#599376) Homepage
    Here at Finland, radionet [radionet.fi] and Saunalahti [saunalahti.fi] are busy building wavelan network. While the idea is great, at the moment they have many techincal problems.

    In practice you need a line of sight and radio frequency is used by many other devices. A common joke is to call it instead of a "wireless Network" a "connectionless network". Still those few who have managed to get a subscription seem to be more than happy.

  • you are kerrrayzee man.
  • well I suppose every luser with a linux box on the wireless network could route a little bit of the traffic through their cable connection. And seeing you are using IP Masq there's really no way the cable company could figure it out. With enough intelligent routing the system could approach an almost completely wireless net with the ISP's just doing the backbone movement between local wireless WANs. Although the wireless connections would want to get a lot faster. Ping times of 2.2ms for each jump is pretty slow.

  • Uh Oh (Score:3)

    by max99ted (192208) on Sunday November 26 2000, @10:43PM (#599379)

    Julian Priest is walking east down Clink Street away from his office. He's holding his laptop in both hands and surfing the Web as he goes through an enviable five-megabits-per-second link to his desktop computer.

    I can see the headlines....

    LONDON - Chalk two more victims for Clink Street. In a continuing display of brutal idiocy, two men were killed earlier today when they inadvertently stumbled onto the busy thoroughfare. Ralph Foster, of York, was "surfing the web" on his new laptop while out for his lunch time stroll. Witness say he had just logged on to Hotmail.com and was happily deleting spam when he tripped near the curb and tumbled out headfirst into traffic. Police say he was killed almost instantly - managing, however, to log out before logging off.

    Simon Edwin, of Chester, had a slightly different fate. With two Palm V's in each hand, Simon was actually playing a modified version of pong online - against himself! The police cannot confirm his exact cause of death, saying only that they have yet to find the 2nd PalmPilot.

    These unfortunate events come just days after the loss of the entire development team from WebSolutions.com, who met untimely deaths when their Internet-connected coach flipped four times after the driver veered to avoid packet loss...

  • The problem with this is, if you want to network with seven other computers you may need seven unidirectional antennas-- for continuous connection.

    the alternative is to setup programmable "antennaebots" that are on a predtermined schedule to re aim the antennae(s) from one computer to the next either in a cycle or to send a "yoo hoo over here" signal through narrowband dialup to tell the other computer you need to connect.

    I had considered this as a possible way to share mp3s that might be better than napster/gnutella in highly populated and geographically flat areas.

    Problem: part of the routing data would probably need to include geographical location making it EASIER to track you to your location. one answer may be crypto/steganography, but even then for it to work people would need to be able to identify files on your computer.

    even with the potential privacy snags the prospects of a network who's infrastructure is owned by the public not by megalithic corporations.

    only with a completely unownable system available can we hold on to what freedom we have on the currently propriatary network infrastructure.

  • 1km?? (Score:3)

    by demon-cw (162676) on Sunday November 26 2000, @10:52PM (#599383) Homepage
    Nice, to start with:-)

    Currently we have a 802.11 based system which can do up to 30km on a p-t-p link and 5 km in radius!!

    We're using a selfmade module for enhancing the signal on the receivers end and 1,8m parabol antennas.

  • ssss;;;;jjjjjjmeeeeing it outsid;;;;ow, ans I jove it!!! ;ometimes I get a little ;;ne nojse, which can be distracting, but over all it's just ;;;;;j jjjjj;l;jjsssssaaaajd jjjjj;l
    Of course, I wouldn't do anything confidential this way... would you?
  • If someone could come up with a peer-to-peer bandwidth hopping scheme, this would actually be pretty cool.

    My guess is that the easiest way would be to build in some sort of GPS ability (so you know where you are in relation to your nearest land-node). Your packets hop from wireless node to wireless node until they hit the land-node. Same technique for receiving packets.

  • Re:Eavesdropping (Score:3)

    by ipl31 (252950) on Sunday November 26 2000, @11:15PM (#599387) Homepage
    I usually dont like to make "smart-ass" remarks, but apparently all your data is based on what you have "heard" and "read", how long ago did you read that magazine article? All 802.11b compliant devices being made today support WEP which a form of encryption used for wirless lan technology (note: some older 80211b cards do not have wep ala AIRONET (before cisco bought them). Also on the other hand here is something to think about with wireless links, you would put an unprotected computer on the net? Well then dont put one on a wireless link, use a vpn over the wireless if you feel insecure. But "stay away from these things" becuase of what you hear is not the way I live my life. Later

  • Seattlewireless.net (Score:3)

    by ipl31 (252950) on Sunday November 26 2000, @11:22PM (#599389) Homepage
    Any one interested in this technology should check out Seattlewireless.net [seattlewireless.net] we are a group building a free wireless network in the seattle area. Along the same lines as consume and other projects, however our mission is not based on internet access, but to create a network unto its own based on 80211b technology. We plan to have internet gateways present on our network, however we would like to see a local/free/public wireless network that had its own web its own irc, quake servers etc... Check it out for more info.
  • The problem is that there are only a few 802.11 bands (3, IIRC). My rooftop antenna is currently on order. But even without that, I've noticed that my in-home wireless LAN takes a hit when my fellow professor, who already has an antenna, does heavy things. We both take hits from people with wireless phones in the same band.

    With a wired Ethernet, if the load gets too heavy we just split the subnet. With my DSL line, if my neighbor gets DSL I don't suffer. But I really don't want my neighbors to get 802.11 installations, because then my performance will drop -- both in-home and (once the rooftop arrives and I drop DSL) to the Internet in general.

    This a fundamental problem with wireless. Highly directional antennas help, but it's expensive to narrow the beam to 1 house from a km away, and so our current installation has an omnidirectional antenna at the base.

  • Heh, and then just imagine that ethernet started out as a wireless radio packet switching protocol on Hawai. Coax was considered an improvement... See: What is the internet [uky.edu]
    The Ethernet concept arose because a researcher from Xerox PARC spent a sabbatical period at the University of Hawaii and noticed that the random access radio system could be operated on a coaxial cable at data rates thousands of times faster than could be accomplished through the air.
    How about using glass fiber to the home instead? Maybe even use wired routers switches to get our packets on the fastest network?!? Then all get what we want and we can reserve the limited bandwidth for radio packets for applications that really need to be mobile, instead of as a nifty techie replacement for DSL.

    --

  • Low Cost Wireless How-To (Score:2)

    by Anonymous Coward
    You'll find the Low Cost Wireless Network How-To [gbonline.com] to be very informative in setting up your own underground wireless network.
  • Sorry it's propertary and it's a little electronic device (no open source code here).
    But you can find further information under http://www.funklantechnik.de (in german).
  • Who do not use ssh anyway? This neither less, nor more, public than the Internet. It can be eavesdropped? Yes. So can my connection from France (where I live) to sweden (where my webserver, IRC-client and more or less everything but my physical body, lives). But any eavesdropper won't be able to _use_ the collected data for much, except as random-source :) That is, if he doesn't have a big numbercruncher and can break my ssh...
  • The problem is defining how the addressing and routing will work - every node will need to be running a routing protocol at least (something like RIP should be OK for a small network).

    The nodes which have a wired connection to the Internet should advertise a default route with a suitably low metric - as these routes float through the RIP network, the one that 'wins' at any node is the route that represents the fewest hops to the wired Internet connection.

    There are probably much slicker ways of doing this but they'd also be more complex to set up.

    Probably all the wired-to-wireless gateways will need to do NAT. They should also do IPSec for those nodes that don't have WEP type encryption.
  • That's actually link layer encryption - end to end means host to host, e.g. IPSec transport mode.
  • I have seen many grassroots level projects for wireless lan in US, but they never succeed. It would be nice if this can take off, but it seems that what pushes the limit of technology is lack of it. Thus in US where we have DSL and cable modem, most people are relaxed, where as in other countries where high speed link is rare and very expensive, people are getting creative to overcome their problems.
  • Is there web site, or IRC channel, or InternetBBS somewhere that intends to connect the different groups building these WLANs?

    Would anyone from Windsor, Ontario, Canada be interested in building such a network?

  • Yes. IEEE 802.11b is a standard that all manufacturers are (supposed) to adhere to. Check out The official 802.11b (WiFi) site [wi-fi.org] for a list of certified compatible hardware.
  • Just hope that the quad, or your pool, doesn't have a dozen 802.11 beams intersecting there. The beams may not interfere with each other, but your radio will have trouble hearing through them.
  • WEP crypto uses the same key to join the network as it does to encrypt traffic on the network. All users using any particular access point can see each other's traffic.

    There is 64/128 bit crypto, in each case 24 bits are taken for the node address, leaving 40/104 bits for encryption keys.

    Crypto is done using RC4, a system with known, exploitable weaknesses.

    No Wires Needed [nwn.com] offers cards with factory-installed public key encryption and Diffie-Hellman support, but they do not offer their cars in the USA.

  • The November issue had a good overview of current wireless 11Mbps products and their limitations.

    Available at http://www.networkmagazi ne. com/article/NMG20001106S0004 [networkmagazine.com].

  • Not to sound snarky... but so what?

    Many people, for several years, have been using wireless LAN cards and doing links of up to 15 kilometers. Several companies even originally based their wireless bridge products on wireless lan cards + some kind of RTOS on a small board.

    I have 802.11 links working over 15Km........ it's *easy*. This is barely even a hack these days, and it's nothing new. I mean, oh my god! You put a different antennae on it, and the radiation pattern turns directional, increasing range! WOW! It's not like every hammie in the WORLD doesn't already know that...
  • My wavelan gold card most definately has a button for 'encryption' as well as some other stuff to do with only allowing other cards with several ID bits set the same to connect.

    Encryption is the key, though..

    *OLD* wlans people, had no other layering to provide security. New stuf most DEFINATLEY does.
  • Umm.
    Who said anything about IP? Many things in windows networkign are discovered by *broadcast*, and use *other* protocols... like ipx or netbeui.

    And by 'log-on' he probably meant he could simply see all their public shares; lots of older networks, especially windows ones, did not use any kind of client authentication.
  • www.waverider.com

    I can say from experience, these are *just fine* for gaming. I don't know where you get your 'the latency is high' figures... but if the appropriate radio mac layer is used, latency is basically no different than a wired network of equivalent throughput. Why would it be slower?

    The *tend* to be slightly higher in latency (measurable, but not noticeable), due to overhead in dealing with the rf medium.

  • You need do no such thing.

    You have to a) obey regulatoroy radiation requirements (dictates power and shape of radiation pattern allowed) and b) use the appropriate antennae.

    For instance: the yagi's we use to do some 10km links radiate a 30 degree wide pattern. This covers *quite a bit* of ground at 10km That means, yes, that a bunch of end users can share a single radio channel, and you only need one antennae each.

    This problem is fundamentally no different than cellular problems: it's all about frequency re-use. You use alternating polarizations, proper channel separation, and can quickly built a tower that has, say, yagis every 15 degrees around the outside, covering an entire circle, giving you some nice coverage.

    Privacy snages? The Internet is alreayd a public medium.. you can't control your packets once they leave your network.
    wireless network cards almost all have built in encryption now.
    Routing data does *not* need to hold your geographic location; it's very similar to ethernet.
  • This is an honest question...

    Do you know whether or not this violates the regulations?

    I know that we have easily done large links like this before, but never within the bounds of the regulations on the ISM bands... I'm sure Germany follows such regulations as well..

  • The protocol was 'aloha', and is used when you have hidden nodes.

    You can't do collision detection, and you can't do carrier sense. Only one central node is guaranteed to see everybody. So they have a system of requesting a slot, etc...

  • DSSS doesn't 'send the same data at the same frequency'. It spreads it over the entire band, constantly (as opposed to a narrow transmission hopping around in side the band).

    I think the answer has more to do with power levels confusing the dsss receivers.

    You are right about which one wins the war though...

  • Before you try adding directive antennas to you 802.11 box, I'd suggest checking if what you are trying to do is legal.

    The idea of adding a directive antenna to a system to get most gain and thus a longer transmit distance is sound. The problem is that you can also interfer with anyone else system who happens to fall along the axis of the directive antenna. This is why most countries regulate these systems on their EIRP, or effective radiated power and not their actual power. The difference between EIRP and power is that EIRP inlcudes the effect of the gain of the antenna. Thus if you use a more directive antenna than the manufacturer installed, you are almost certain to be illegal.

    D.


  • Shamelessly lifted from BAWUG [bawug.org]'s links [playanet.org] page, where there is lots of information about wireless hardware and software:

  • In an improperly configured wireless network, yes it is possible to eavesdrop. But so can your home telephone line with an op-amp tool (the ones the bell guys carry). So can your cordless phone.

    When a wireless network is properly planned, there are 3, sometimes 4 or 5 layers of network security:

    1. In freqency hopping systems, there are 78 hopping patterns. Only one will "sniff" the packets in the correct order.

    2. ESSID. This is the network security password. Access Points will only respond to radios with correct ESSIDs.

    3. WEP. Wireless Equivelency Protocol. This is 40, 56 or 128 bit encryption that encrypts all communication on the radio.

    4. Direct AP connection to router/firewall. This stops any extraneous traffic from flowing over the WLAN.

    5. VPN Encryption. Many ISPs are using VPN from customer premise gear to the router, which makes sniffing wireless more difficult than wireline, since there are 4 extra layers of security.

    ---

    No security will ever be perfect, but then again, most people using wireless lans are using them for internet access which is inherently PUBLIC anyways.

  • Re:Eavesdropping (Score:3)

    by GC (19160) <giles@coochey.net> on Monday November 27 2000, @10:50AM (#599462)
    We've implemented a Wireless link with Encryption to connect multiple sites. I have to say that I'm impressed with the performance.

    To join the network you have to be a registered device on that network.

    To eavesdrop you have to break the encryption key.

    The alternative to wireless was to implement a fibre link, but it seems to me that it would be easier for someone to dig into the ground and plant a device on the fibre much easier than attempting to break the encryption keys on a wireless link.

    Note this:

    With our wireless link there are a fair few checksum errors on packets, resulting in the odd retry request from the other node. I presume that these errors would not be known unless you knew the encryption keys, making it almost impossible to crack the keys as you cannot tell where the errors are without first knowing the keys. Has anyone done any research in random (deliberate) corruption to encrypted traffic in order to prevent key-cracking? Obviously this would come at a performance cost (extra re-transmit packets), but I guess this is to be expected for higher security.

    I'm sure anyone with enough spare time could probably break the keys, but probably not before we changed them.

  • Re:Eavesdropping (Score:3)

    by Zppr (22841) on Monday November 27 2000, @11:54AM (#599464)
    Carnegie Mellon has a decent sized Wireless LAN [cmu.edu].

    To obtain a lease from the DHCP servers you have to be a registered [cmu.edu] device on the network.

    This approach seems to work well enough along with secured client applications (AFS, IMAP, etc via Kerberos [kclient]).
  • We run high ground (mountain top) Aironet BR500 repeaters/AccessPoints in the US which deliver about 30-31dBm of EIRP (well below the FCC max) into omni antennas and 4800 series radios with 24dBi dishes at the remote ends. We have stable links in the 16-18 mile range, and have tested out to 25miles. The hill tops have co-located high power transmitters on the sites (TV, FM, Cell Phone, paging, mobile radio repeaters, and the like) which contribute a significant amount of broadband noise to the site. As a result FHSS perform much worse than DSSS systems (which are able to recover signals below the noise floor). Our repeaters are therefor receiver noise limited, which limits our range even though we have the remote power tuned right to the FCC max on the dishes. In theory, this stuff can work out to the horizon (65-100 miles) if the environment around the site is RF clean (very remote) - and in dicussions with the military boys it does just that in certain areas. Even so, with three hill top repeaters we have better than 30% coverage of an area 40 miles wide and 55 miles long (just over 2,000 square miles under the antenna pattern, with an effective coverage area of just less than a 1,000 square miles after subtracting out shadowed areas). We will probably double that during the next year. Our market is NON-Wired high speed services, and we suggest potential customers that can get wired access do so. Some of our customers live off the grid - no wired power/phones - most do not have access to ISDN, DSL, Cable, or T-1's. We are a member owned, member-operated cooperative, with the goal of providing ourselves the service we need (when AT&T and Qwest cann't, or won't).

    We are a coop for a reason - it was clear from the beginning that you can only build out *ONE* of these wireless networks in a region. We allow equal access to our network by all regional ISP's that want to particpate for a $3K startup (includes radio/routers which are coop owned/managed) and $300/mo. In short, a turnkey service far less costly than the 2 man years and $25K it would take to rebuild the nework form scratch. Coop members pay $60/mo for service - about the same as DSL and Cable service in this area - and lot's cheaper than ISDN. We are not a free open source like entitiy ... but as a member-owned, member operated cooperative, the next best thing.

    Making it work more than a few miles without a 2500ft tower (aka mountain) is very difficult. Signal (wave front) diffraction is driven by pure physics ... and every object the signal passes bleeds energy off the wave front to fill in behind the object. So you lose 3-6dB over every building roof and tree that you barely clear with LOS. The diffracted wave front also reflects out of phase to distort the main wave front (multi-path) - creating a noisy/choppy signal.

    So flat landers with lots of buildings and trees, and no serious height, pretty much will have to live with short connections (under a mile or so), probably much less in typical city environments. This is much less a problem at 900mhz, is difficult at 2.4GHz, and a total killer at 5.7Ghz.

    While some vendors say that 2.4GHz isn't affected by weather, that is only partially true, and only for links that have 15-25dB or more of link margin. Many of our 2.4GHz links have less than 10dB of link margin, and see slight rain fade, but serious signal loss due to snow which causes serious diffraction problems coupled with broadband noise refection problems at the repeater sites (noise floor goes way up during snow storms). We manage this by dropping the modulation rate to 5.5mbps (doubles the power per bit) and decreasing the packet size (enabling fragmentation) to significantly reduce the probability of CRC errors due to noise functions. At 5.7GHz rain drops just completely eat the signal, and rain fade is really rain-block.

    Network performance with Hidden node operation, while managed by 802.11b, degrades rapidly underload. Aironet made a huge mistake when they failed to implement the Point Coordination Function called for in the 802.11 spec (PCF). In theory PCF can be used to stabilize the load curve, minimize load induced failures/overruns, which are a fact of life with hidden node architectures when using 802.11 devices in a wide area network.

    There is a lot of RF magic in making 2.4GHz 802.11 wide area networks work - the stuff is worse than plug-and-pray. It pretty much takes a $2-20K spectrum analyzer investment to debug problems. Even with that expect side by side experiments to have radically different results.

    Watchout for mixing FHSS and DSSS systems in the same area ... they do not mix well in weak signal applications. In close applications, the 802.11 spec requires them to check for energy in the channel and hold-off transmissions to minimize collisions. In wide area applications with all hidden nodes, they do not sense each other, and collisions rapidly degrade to inoperation.

    Or operation isn't perfect ... we do see brief periods of in-band interference, and high-power broadband interference which causes several second to several minute drop-outs in our service. Sometimes several per week, sometimes several per hour. But compared to unstable 19.2kbps dialup on rural phone lines ... it is completely heaven to get megabit web surfing with dedicated connections.
  • The November issue had a good overview of current wireless 11Mbps products and their limitations. Available at networkmagazine [networkmagazine.com]

All trademarks and copyrights on this page are owned by their respective owners. Comments are owned by the Poster. The Rest © 1997-2009 Geeknet, Inc.