Slashdot Mirror
More Wireless Networking for Linux
I was sent word that AbsoluteValue and Intersil have announced more development of wireless support for Linux. The announcement is actually for Intersil PRISM-based WLAN cards, but what it does mean is more support for IEEE 802.11 wireless cards, access points, etc. Here at Slashdot, we've been using the ZoomAir cards, but are starting to play with the Lucent cards - what does everyone else think about Linux and wireless cards?
We've been using the Lucent WaveLAN 11Mbps cards here, and they work like a charm. Lucent supplies (binary) drivers for their cards which work quite well. The only annoyance is that the access points have to be configured remotely VIA M$ Windoze, but it all works really well. In addition, they interoperate with the Apple iBook/AirPort stuff, too, IIRC.
The Aviator cards from WebGear are very good. And fairly inexpensive as well. You can get a 2-station kit here. I use them at home and can take my laptop to a friends house and walk right into their network by bouncing dhcpcd. I get about 1.5Mbit inside my house too, which isn't bad considering my internet connection (also wireless) is only 256K.
Ive been happy with the Webgear (Av 2Gz) wireless lan cards. They support Linux and I havnt had any problems with them. You can get 2 PCMCIA Wireless cards and 2 PCMCIA-to-ISA cards all for $140. Range is about 1000 feet outside and 500 inside. I think their site is www.webgear.com.
I have to return some videotapes...
The Sierra Wireless Aircard is a good card for laptops running Linux, they've even got a link on how to get it running under Linux on their web page.
I have zoomair wireless networking set up both at home and at work and I love it. nothing like going out to the pool to hack on a nice day.
--
Geoff Harrison (http://mandrake.net)
Senior Software Engineer - VA Linux Labs (http://www.valinux.com)
Geoff "Mandrake" Harrison
Some Random UI Hacker
I've got a Proxim Symphony network that works across all of my family's houses and I was pleasantly surprised at how easy it was to get my Sony VAIO 505 to run the card under Linux. Binary module is available and with the exception of APM which screws up the card when you suspend/resume the laptop it works as well if not better than W95.
there is a low cost wireless how-to at www.qsl.net/n9zia
I've been using Aviator Webgear 2.4GHz IEEE 802.11 cards for a while. You can buy a 2-pack for only $139.99 at the nearest CompUSA; and the kit includes 2 ISA adapters as well as the two PCMCIA cards. They work great with Linux, too. I recommend Aviator 2.4 cards to everyone instead of proprietary stuff like Diamond HomeFree; which do not work with Linux at all(well not really, but the drivers work in a non-FCC approved way and can't talk to Windows drivers)
Now that we have affordable IEEE 802.11 cards that work with Linux, I guess the next step is to have IEEE 802.11 access point capability for Linux. I guess having an access point is the easiest way to bridge seamlessly between an Ethernet and wireless; I did not have the time to deal with finding an alternative solution; so my wireless LAN is just another segment with its own non-routable class C; and I use masquerading at the gateway. I've been checking out the Absolute Value site for a while, but they seem to be concentrating on cards with the Intersil chipset.
--
BluetoothCentral.com
A site for everything Bluetooth. Coming soon.
Zigbee Central: A Zigbee weblog
Check out my post about the Aviator WebGear 2.4, which is IEEE 802.11 compatible (2Mb/s). At $139.99 for a 2-card kit (at CompUSA), it is a very good value. Works with Linux like a charm, problem-free (on both laptops on desktops with the included ISA PCMCIA adapters).
Once you go wireless, there really is no going back.
--
BluetoothCentral.com
A site for everything Bluetooth. Coming soon.
Zigbee Central: A Zigbee weblog
We're using the Lucent WaveLan stuff here at CMU.
I've only used it under windows (no real good reason for my vaio to be running linux), but it's quite nice and performs really well. The software that comes with it is really cool -- you can see graphs of the network reliability, remember other users by their MAC address so you can see who else is using it near you, etc. We're not running encryption right now, which is somewhat tragic, but the WaveLan silver stuff apparently supports 56-bit DES. I've heard that it works well under linux, and have seen many machines running it.
We sell the breezecom stuff at work -- it's more expensive but boasts some nice features. They claim they can get 5 miles out of one station adapter -- which would be enough for me to connect to work's T1 from home! No idea about the linux drivers situation for this stuff.
Lucent makes the nicest-looking wireless equipment that I've seen, too (no ugly antennas!). Breezecom's hardware appears very amateurish (reminds me of a Commodore 64 or something).
- Tom 7
As scott__ said to me today, having a web pad is interesting, but having a wireless modem for a Linux-based Transmeta chip web pad is really interesting.
This is something that people will buy for their homes, as well as for their offices. When a car dealer can carry around the pad and order it for you on the lot - that's web-enabled data entry. When your doctor can run the web front end for the medical database in the secure internal wireless LAN in the clinic and update your data while giving you a check up - that's interesting.
When I can undock the web pad from the fridge or TV holder and start the bath, turn on the outside lights, check the web cam from my son's room, look at his homework chart from the school (darn, he has one of those silly papers due!), see that my Cisco stock is at $562 a share at the limit I set to sell some, sell the stock in the aftermarket, transfer the funds to my credit card, set Expedia for Linux searching for cheap fares to Hawaii, and cook dinner - now you're talking!
Will in Seattle
I have worked with wireless, I have used the Breezecom, Lucent WaveLAN, and Solectek Corp. wireless, and I am happy with the results, in so far as a solution for a remote site, but when you are dealing with 900 mhz, the licensed MMDS band (2.500 to 2.690 GHz), or unlicensed U-NII band (5.725 to 5.825 GHz). You have to realize that those are all microwave, or in the case of the 900mhz almost microwave frequencies. Now I may be paranoid but a microwave has heavy layers of sheilding that stop the propagation of radiation, but these things have antennas that help propogate radiation and you stick a card in your laptop and set it on your lap. I don't know. It's just to creapy for me, also the latency of microwave wireless over a relitivly short distance is far to much to get any good voice over ip, or do anything that is time sensitive, I.E. q3 and such. It's a good choice for some instances, like browsing the net, chatting, or e-mail, but don't think it can give you everything that your lan can.
and I love em. I detailed how they're set up on our silly little YALP on the webpage. I've been using the kernel bridging to connect to the LAN because I dont want to route YET another subnet. Only problem is I'm getting very very slow speeds across the bridge. I was just told today to try a replacement bridge, so I'll give that a shot. While 802.11 access points would be nice, bridging should be sufficient (at least for me).
I did a survey from spring 99 that was aimed at robotics use. At that time I saw it as a good possibility. It can only have improved sense then. I eventually plan to use them for linking to my robot and a wearable computer.
Wireless is coming in a big way. First 802.11 got stable, then supported by major chip vendors, then Apple announced integration in the iBook. I suspect by the end of the year it will be one of the things you could never had dreamed of living without. The ability to walk into any airport and many hotels in the US and get cheap wireless net will help get it out there.
Personally, at work I use an Aironet (now owned by Cisco) 4800 card (under linux) and have had no problems. Some chorts of mine are currently setting up a point to point wireless conection with yaggis on the end of the same wireless cards in PCI carriers - seems the local carrier couldn't deliver a T1 fast enough, but there was no problem getting roof access...
Expect to see 11Mbps cards in the ~$100 price range soonish, with slightlylower prices for 2Mbps versions (based on older chipsets)
Around the same time, expect to see 'cable-modem-top' boxes that integrate a wireless bridge into a hub that you can plug your cable modem into in order to get portable net in your house.
All of this is for 100meterish-range wireless.
For closer range expect to see Bluetooth moving into all the places IRDA tried to go with only marginal success. Walkup printing, PDA access, even wireless headphones. It's too cheap and robust to fail.
For longer ranges, well, there's some work going on to use the cellular network for packet data, but here big companies and big money get in the way of the quick innovation that is needed for technologies to be tried fast enough to find a workable solution.
--Xandir
About the same amount it costs to make anything else of the same size.
The cost is based on R&D, and the fact that they don't sell nearly as many. When everyone is buying wireless network cards, they'll be dirt cheap.
Mainstream? I don't think that wireless cards were ever targeted at mainstream users.
That's funny. You're implying that 10/100 cards and cables are targeted at mainstream. I think the opposite is true - wireless is targeted more at mainstream than cables ever were - or it will be soon.
Anyone remember when 10/100's cost > $100? How long do you think this will take - especially with Apple pushing the Airports...
Having used both Proxim Rangelan II, and several versions of Lucent WaveLan (all pcmcia cards) in linux, I didn't have to many problems.
Neither has full source, you need a binary-only module (or in the case of the proxim driver, a library that you link against, I believe built from the Proxim reference source.) The Proxim driver was a bit bitchy in it's early days, but I'm told it works very well now.
Another note... most wireless LAN solutions have some kind of 'base station' or 'access point' they expect you to buy. If all you have is a few PC's, and it's for the home hacker, you generally do not need this. The cards themselves can communicate with each other just fine. The base station usually serves as the bridge (possibly router) between your wired and wireless networks, sometimes with added features like supporitng multiple transcievers on multiple channels.
Another trick with wireless LAN cards.. if you replace the omnidirectional antenna they give you with a good directional antenna, you can massively extend the range. I have seen Proxim Rangelan II cards bridge distances of about 15 Km and still be within regulations. This also works if you have a wall you need to penetrate that you can't quite manage without. (depends on the materials, though..).
Also.. most wireless LAN cards, by regulation, must use a non-standard connector. Usually, it's a standard connector with the mating parts switched around. (this is to prevent joe average from hookingn his little card up to his linear amplifier and causing WW-III).
Hmm. I could go on for a while.. perhaps I should start a Wireless LAN HAcking FAQ?
My local ISP wants to experiment with a wireless connection. I'm looking for something with a 3 or 4 mile range. Does anyone know of any good solutions for under a grand?
>Nobody has said anything here (yet) about the security of wireless LAN cards. If I run a tape backup across my LAN can someone sit outside my house and scoop it all up?
Pretty much. Minor security in the form of an SSID that will keep the naive user from connecting to your net (it's really a way to run multiple wireless subnets that overlap), but nothing to stop the sophisticated user. Tcpdump works fine. So SSH is your friend. But of course the Men In Black have Tempest technology.
--Z
Here at work, we recently (two weeks ago) implemented a four-block wireless shot direct from our main office, and it works like a charm. We used two Linux machines as the routers, with the PCMCIA-cs modules loaded along with the "wlan2" module which is available from Lucent here. The WaveLAN "silver" 2.4mbit cards were used on an ISA-PCMCIA bridge.
We got a professional communications company to run up antennas on towers on both sides of the shot, and it's great, no problems at all. We commonly get 5ms roundtrip times, and can push well over 1.6mbit (200Kb/sec), 0% packet loss, and the weather makes no difference from what we can see (and we've been through some heavy snowfall and rain already). As for the technical setup, in all honesty, any competent sysadmin should be able to set up such a connection in a snap -- it's that easy.
If it's a long outdoor wireless shot (as opposed to wireless for an indoors network only), you might be able to go out and climb onto a tower and do it on your own, but there are companies which exist to do that (antenna installation, pointing, etc) already. Then again, if you're trying to save cash, and want to hook up your friend a block or two away into your cable (or *DSL) connection via masq or whatever, then I suppose you could do that :)
However, it's not like this stuff is cheap, the cards are commonly $200cdn or so, with the antennas being just as much. Low-loss cable for the antenna is expensive, too. There are a lot of things to factor in, but all in all, it's all pretty cool stuff :)
It's great to see more and more companies supporting wireless on Linux, though, as it means not only will we get more robust drivers, but installation will be easier, and support will be broader. Of course, it should be obvious that anyone using wireless for the same thing we did (i.e. linking up two networks across a few blocks, or even kilometers away), they'll be using anything but Win* to do the routing work for the greatest amount of stability.
Wireless is insanely useful for the portable market. Portable devices that aren't tethered by a cable are becoming more and more prominent in the workplace. The cost however is somewhat justified (note that I said somewhat - read below).
There is a lot of processing power in those little cards to implement the IEEE 802.11 spread spectrum model. They're not cheap for that very reason. Plus you don't have as huge a cable infrastructure, which is a cost most places forget about.
There are on average about 3-4 times as many components/transistors on a wireless card as there are on a comparable network card. You also have to remember that almost all PCI/ISA wireless cards are really a PCMCIA card in a PCMCIA adapter. Compare the cost of a PCMCIA network card to a PCMCIA wireless card and the difference isn't as great as it first seemed.
Why stick with PCMCIA and use adaptors? Well for a start, developing 2 cards in completely different technologies tends to take way too much time, and leads to revision problems. Add to this that the greatest market for wireless is actually portable devices that have PCMCIA slots, and you can see the economics that cause the choice.
However, I'd still like to see a Compact Flash version of these (and using a CF to PCMCIA adaptor, you can still stick with one version), even if it sticks out of the Compact Flash slot. This would bring wireless solutions to devices that could really benefit them. e.g. Cameras - you take a pic, it stores it locally, and then sends it over the wireless link to a box nearby to store it. PDA's - allows you to access network resouces when in range. And so on.
There are also H.323 compliant phones available that use IEEE 802.11 spread spectrum. They operate across your normal wireless network to provide wireless telephony. Now if only the cost would come down, I'd be happy! *sigh*
The range can be extended greatly if you use a directional antenna. (one on each end) I have done this, and it works like a charm.
THe ratings they give are for their little crap omnidirectional antennas, that have almost no gain.
A directional yagi will give you great directonal gain, and allow you to do what you want. If oyu aren't comfortable with it, explain it to a radio technician who can make the appropriate custom cables for you.
To make an example... one pair of cards, rated at 600 feet, with the appropriate antennaes turned into a 17Km link from a mountain top to a cabin. Not bad for a few hundred bucks in wireless lan cards, and some good radio antennaes.
Also.. running cat5 cable between buildings is not such a good idea, you run the risk of ground loops and burnt wiring. Better to use 10BaseFL (cheap plastic optical cable) if you want to do it this way.
Anyone who knows anything about this technology knows what Im talking about. First of all, the price. The prices are around $30 for the chipset but it is expected to drop to about $5-$10 once mass production ramps up. This means they will show up everywhere.
So whats so great about bluetooth? Well it was orginally designed to replace irda but it will now be doing so much more. Bluetooth devices have about a 30ft range. When ever one device comes in contact with another, they set up somewhat of a mini network. Eight devices in all can make up one of these networks. One is the master telling the others what frequencies to skip to ect (Bluetooth operates at 2.4ghz and utalizes spread spectrum much like digital phones). These networks can then communicate with other networks of devices. This is all done on the fly and may be reconfigured constantly when new devices move in and out of the area.
So what can be done with this. Well, first the obvious, like networking printers ect but it has so much more potential. From what I hear, the bandwidth is comaprable to dsl which is plenty for most applications and other bluetooth devices dont hog the bandwidth. Well, for one you could could dial up a connection using your phone in your pocket without even touching it. Now imagine bluetooth everywhere and commercial services. Imagine negotiating for an internet connection in an airport without ahving to talk to anyone. Imagine pulic printers, and a slew of other things.
We are only beginning to realize what wireless technology can do. The potential is gigantic. There are some things to worry about like will each device have an id; autonegotiation is scary. 2.4ghz is also the resonate frequency of a water molecule, think leaky microwave. Still, I believe this is going to be amazing, 1) because its so cheap and 2) because its so easy. Truely plug and play networking. Expect to see bluetooth devices this spring and integration into laptops and phones by next fall. For more information, check out Bluetooths home page. Lots and lots of big name backers. This is no vaporware.
Since this is being discussed, I'd like to mention that I've had great success using a Lucent WaveLAN silver at LISA '99, and subsequently bought a Lucent WaveLAN Gold and Apple AirPort Base Station for use with my subnotebook at home.
I wrote an Enlightenment Epplet to graphically monitor the signal strength to the card, something that comes in very handy if you're wandering from room to room or building to building.
It's called E-Wireless, and is available from FreshMeat here.
I haven't yet adapted it to WindowMaker, Afterstep, and such, but it shouldn't be difficult at all.
I'm also currently working on an app that will allow you to set up an Apple AirPort Base Station from Linux.
.@.
here's the URL of the Apple AirPort doco that compares DSSS channel availability in different parts of the world
5 50
http://til.info.apple.com/techinfo.nsf/artnum/n58
brian whitaker
applications engineer, maxim
bwhitake@knarf.com
The only issue (besides cost of course) keeping me from currently switching over to wireless networking at my place is security concerns.
Can anyone tell me about the potential security issues that arise once one is literally broadcasting their information? Is (for example) SSH just as secure over the airwaves as through an ethernet medium?
I have a 100baseTX connection to all of my boxes throughout my apartment, and that more than sufficiently serves my purposes, but it'd be nice to take my laptop outside to relax from time to time...
--Cycon
Your Brain + EEG + LEGO Robots = Brainstorms
Bluetooth products are about to show up pretty soon, and I agree it will be a hugely successful technology for things like you listed. Since the eventual target price for Bluetooth is $5/node; if they manage to meet that price target in two or three years, we might have $10 Bluetooth LAN cards that work for very short range (~50ft) at a rate acceptable for basic tasks (741 Kb/s).
Do you think Bluetooth might be successful in this role? It is definitely not something Bluetooth is designed for; but it will be so inexpensive that inevitably this will be an application.
--
BluetoothCentral.com
A site for everything Bluetooth. Coming soon.
Zigbee Central: A Zigbee weblog
I could afford wireless networking cards for all my boxes. *looks under his desk at all the cat-5 waiting to get tangled in the wheels of his chair* Oh boy don't I. I've wired half my house with cat-5. Wireless networking would be really cool especially since I could hang out on the back padio on my laptop reading /. but it is way too expensive for me. Enough cat-5 to wire my house and 4 NICs == 50$. One wireless networking card == 50$. It would be kinda nice having an AirPort hooked to my cable with all my other comps hooked wirelessly to it. So-called broadband available to the you and mes of America wouldn't max out 802.11a's bandwidth so it would be great for my house.
I'm a loner Dottie, a Rebel.
The real killerapp for wireless tech will be a bridge, of all things - just a plain old transciever that plugs into any hub, switch, whatever. To date I have found only one such device from Proxim, and it costs over a thousand dollars. Think about it - no interplatform compatiblity issues, true interoperation with existing infrastructure, and no taking up slots in your server for radio equipment that could and should reside outside of it. I can't believe Diamond et al. didn't think of this when they made their home networking push. Carrying a signal over the phone lines is great and really, really convenient - but I am not going to downgrade from 10mbps Ethernet on 6 of my computers to 1.2mbps phone lines on 7 when I add another PC. If I could buy a HomeFree kit and a bridge from phone line -> ethernet into my existing hub, I'd have bought 10 of them months ago. Instead, I'm forced to sacrifice weekends stringing cable to whatever room gets the next PC (7 and counting ;). The same goes for any of the electric line packages, etc. Give us the bridges!
--
I think there is a world market for maybe five personal web logs.
Why does it matter? Wireless networking xmitters should understand ethernet packets. They should not care what OS is producing them.
And in fact they don't. At least not as far as transmission goes. But once you've caught a packet and stripped off the header, what do you do with the contents? Send 'em out serially? Parallel? How are you marking start and stop? That's where OS (and hardware) dependence comes into the picture. And like it or not, one particular OS has most of the market among the businesses who can afford to buy new technology, so that's the OS that people build for.
Because of the OS independence of ethernet, I believe it would have been a better choice than USB as a peripheral bus.
How do you figure? You want as little management overhead on the data part of your peripheral bus as possible: that's why PCI has so many special signals (REQ#, GNT#, FRAME#, etc.) that are just for control of who's got the bus. This leaves the address/data lines free for shoving data as fast as your clock can mark it off. This is an example of the basic rule that hardware can accomplish a given task faster than software but at greater cost: the PCI bus is some 47 signals wide. But pulling 47-strand wire would be a real pain, so Ethernet makes do with 8 and accomplishes all the overhead by encoding it in the packet header, which could be considered software for the purpose of the rule above.
If packaged ethernet PCI cards can sell new for $8, then the ethernet chipset can't be all that expensive to stat building into printers, scanners, etc.
Perhaps you've never heard of a print server? Or network attached storage? These are examples of peripherals that are on the network. The time delay communicating with them is acceptable because people can deal with printing or file accesses taking a while to complete. But would you want your mouse or keyboard to have the same kind of latency? I think not: you'd be shot down each time you respawned while you waited for your mouse movement to process.
But ethernet is not Intel's baby like USB is, so it'll be build onto mboards to the detriment of all that is better. Firewire never had a chance except maybe in the Mac world.
Since when is USB Intel's baby? Last I heard it was its own standard, as in standards organization... USB is not a competitor against Ethernet, it's a competitor against the serial port (and a vastly superior competitor at that). And Firewire is a competitor against, well, something like AGP and something like the home entertainment network that doesn't presently exist. It achieves its greater throughput by eliminating some of the error checking of USB, so you wouldn't want it for file access, where every bit is critical. But it's great for streaming music or video because a bit here or there on those is no great loss. Ethernet doesn't compete against either one--it's for communication among independent systems, where a peripheral bus is for communication among subparts of one system.
Jenny
I have sold a similar problem (a bunch of Suns on a different floor) by buying an old 486 from a nearby college surplus store for something like $15; and I installed Linux along with an ISA Ethernet and the Aviator on it.
In my case, a 486 SX-25 with 8MB of RAM works flawlessly as an Ethernet-to-wireless bridge. I recommend you get a cheap, old 486 powerful enough to run Linux, do a minimal Linux installation, and either set up kernel bridging or use masquerading to connect your Win98 box to your VAXen or what you have at your home. I have mine on a small UPS; and it just runs away from sight in the depths of a cabinet. The whole thing should cost about $30 with the exception of the Aviator card.
--
BluetoothCentral.com
A site for everything Bluetooth. Coming soon.
Zigbee Central: A Zigbee weblog
Yeah, I'm one of those guys who thinks it'd be Really Cool to have one less cable coming out of the side of his notebook. And I'd like to be able to take the notebook out in the yard, etc.
The Aviator cards sound like a fantastic deal, there's just one thing i wonder about.
How interoperable is this kinda thing? I ask, because my employer has a Lucent WaveLan Bronze bridge set up at work, and it would be Really Slick if i could carry my notebook to work, change the SSID, and hook up there as well.
I like the Lucent gear, sure, but sheesh it's expensive. I really like the price of the Webgear setup, but, are they doomed to be a closed loop? Will i have to buy another card to set up a bridge at work to use the same wireless setup there?
This is just like television, only you can see much further.
TCPDump works fine depending on the card in use.
It depends on the layer 2 protocol used on the radio side. In 802.11, this may be the case.
not all wireless lan cards use 802.11 (and certainly, many that do use 802.11 are not interoperable; 802.11 applies to several bands.).
Proxim, for one.. you can't just sniff on those cards. won't work. The driver (currently) doesn't support it, and I believe the firmware in the proxim radio (that determines how the protocol actually works) doesn't allow it... but proxim doesnt' use a CSMA/CA method.. they use a more detailed protocol where each host has to ask for permission to transmit to it's destination.... and other cards ignore.
That depends on what's in the way, and what frequency you are using.
if it's 900 Mhz and a few trees, you might be okay.
If it's 2.4 Ghz and some trees.. hard to say. might work great.. might not.
If it's 3 new york blocks of steel reinforced concrete buildings.. not a chance.
As far as I can tell, 802.11, by itself, does not guarantee that they all play well together.
802.11 covers infrared, FHSS, DSSS, several bands.
Now. I'm ignorant, I haven't read the whole spec, and i"m positive a subsection of it does deal with exact physical (radio) modulation & the like... (like 802.11a, 802.11b, etc... who knows).
Just as 802.3 by itself does not indicate the media type (is it 10Base2? 10BaseT? 10Broad25? 100BaseFL?).. 802.11 by itself does not specify the radio band or modulation type.
All you can confer when someone says '802.3' or '802.11' is the frame format, and an overall understanding of how the network functions at layers 1&2.
Subsections will indicate how various connection methods work. In 802.3, subsections describe physical media of various types, essentially defining standards. 802.11 will be the same way, but it's the frequency and modulation that will be at issue? is it AM IR? FM IR? DSSS 900Mhz ISM? FHSS 2.4Ghz ISM? (subsonic accoustic?) Gravity waves??
Just as 802.3 does not specify the physical media, 802.11 does not specify frequency/modulation.
802.11 simply describes the protocol to be used of the wireless media, whatever it may be.
802.3 specifies the frame header, the preamble (used to synchronize the receiver), the checksum, the CSMA/CD mechanism, the backoff mechanism, etc...
802.11 does the same thing for wireless. The media is still variable.
In at least 3 of these places, wireless networking is going to be used. Since we're in a mixed environment (Windows, Mac, Linux) we have no option but to use Lucent's cards. (do we?)
... Breezecom also has drivers for those O/S's (they even have ones for the HP Journada)
... sets up connections with 1024bit RSA, then switches to 128bit BlowFish to run. Does cross-subnet browsing for WinBlows clients (SMB/NetBeui)
... you don't want to tunnel crypto to another site and then go open wire because the wrong gateway is set).
... I've seen 128bit run nicely on a P-166, but nothing lower than that yet.
No
- have the ability to surf the net in each of those places.
Just hook up a Breezecom AP-10 to each LAN, set it to "roaming" mode and you're done.
- be able to "see" the pc's on the lans of the distributed locations, as though they were on thesame lan. (VPN)
- use crypto to secure the VPN.
- use a transparent proxy setup to mask the forced use of a proxy for webtraffic.
GateWeaver has a good VPN solution
- have the ability to use the laptops in each of the locations transparently. (3 locations with wireless)
- i'm hesitating to use dhcp on the lan/vpn: to avoid being stranded if a node goes down, all the routers must deal out ip's to their own lans. They have to watch it not to step on each other's toes in the vpn though.
Best suggestion : set each internal LAN to use its own subnet (ie 10.0.1.0, 10.0.2.0, 10.0.3.0, etc), VPN them together and have DHCP assign IP addresses to the laptops (this way they get the right router gateway address
Your 486 boxes will probably not have the horse power for VPN crypto
Wireless Internet in Southern Ontario available NOW! WDSL Inc.
I'd strongly recommend you implement IPSec with 3DES (168 bit keys) between all wireless LAN nodes. I think the WEP encryption has a fairly pathetic keylength (40 bits?) and in any case the usual arguments about open source crypto apply. Even if you use a closed source IPSec on Windows, at least IPSec has been publicly reviewed. Bruce Schneier has reviewed IPSec and was quite critical of its complexity and other problems, but admitted it is the 'least worst' option (see http://www.counterpane.com for his interesting paper on this).
I'd recommend using IPSec's ESP + authentication, and not bothering with AH, which cannot go through NATs or IP Masquerading. You can use ESP tunnels to go right through a NAT and support wired or other VPN access to your home network.
I'd prefer to use tunnel mode even within the LAN, which is also a Schneier recommendation - then you have a clear distinction between encrypted and unencrypted interfaces - if you see any packets between non-tunnelled IP addresses, you know at once they weren't encrypted.
I haven't got this set up at home yet (still waiting for a colleague to dig out his free Nokia WLAN card - they were giving them away at Interop last September) but I will real soon now. I did use BayStack 650 cards at the IETF last year, but these don't have Linux support.
In a sense, all wireless access is dialin, since anyone can just fake a MAC address, stand next to your house and get right in to your network, so strong VPN support is a Good Thing - you might want to make your Wireless LAN gateway sit outside your firewall perhaps.
There are IPSec implementations for Linux, *BSD (check out http://www.freeswan.org/ for Linux and http://www.openbsd.org/ for OpenBSD), and many commercial ones for Windows (I think PGPnet is a free VPN client for Windows but it's only available in US).
One other thing to think about is DSSS vs FHSS - these are two variants of spread spectrum (direct sequence and frequency hopping) - cards can be 802.11 compatible and use either, but the two types cannot communicate... DSSS appears to be winning for 802.11 (2 Mbps version) but there's some questionmark over the allocation of spectrum for 802.11b DSSS (but I could have mis-remembered that last bit).
Wireless LANs are definitely the future - the standards are a huge mess, though. Things to check out include HomeRF (includes SWAP), BRAN (formerly HiperLAN 2, from ETSI), Bluetooth, and more, including the whole mobile phone saga of HSCSD, GPRS, EDGE, UMTS, and so on. But 802.11 seems to have a fairly good niche in Wireless LANs, so it will probably do OK.
Some WLAN vendors are talking to telcos about rolling this out to hotels and airports - the idea is you just walk into the place, sign on to your existing ISP (if they serve that area) or put in a credit card number on a web page, and then you're on the Net at 2 Mbps or higher.
I forgot to say that IPSec can encrypt all IP traffic on your WLAN, just about. Some things, such as broadcasts or multicasts, may have problems, but generally it just works.
You can use all your normal applications, and there's no need for SSH on the LAN.
if you replace the [...] antenna [...] you can massively extend the range.
I'm impressed with 15km !, although I've done this myself on similar kit with good, but lesser, results. My experience was in the UK with MPT1340 pre-built modules at 418MHz (car alarm / door opener tech) - maybe the use of spread spectrum for the wireless LAN accounts for your greater success ? (all hail Hedy)
still be within regulations.
Now this I'm not so sure about. Certainly UK rules for the MPT1340 are very specific about keeping with the standard antennae designs they were certified with (not that it stopped me !). Some of the model rocket fliers and ferret keepers who use these things as retrieval homing beacons have built massive reception Yagi arrays.
Wirles LAN Hack FAQ. Yes Please !
Have a look at http://www.freeswan.org/ - this is a relatively complete implementation of IPSec for Linux - includes IKE for key exchanges but no PKI support yet.
Luckily encryption issues have already been sorted out to implement virtual private networks over the Internet. See my post earlier at http://slashdot.org/comments.pl?sid=00/02/01/16522 10&cid=215 for details of how you might use IPSec for this.
One slightly odd implication is that the WLAN is a public network, so it should really be connected to the outside of your home firewall - otherwise you have to carefully IPSEC every host in your home, even the wired-only ones. Or you make the WLAN gateway into another firewall, but why have two firewalls?
Wow! 2 cards, 2 ISA adapters, $140 at CompUSA, and they even work with NetBSD and Linux.
This looks like a pretty good deal. They're not as good as the higher-end cards, I'm told, but they still seem pretty neat.
Currently setting up my NetBSD gateway machine to be a gateway to one more network...
My blog: http://www.seebs.net/log/ --- My iPhone/iPad app: http://www.seebs.net/seebsfrac/