Slashdot Mirror
Sluggish WiFi Connections Hurt Everyone
MindNumbingOblivion writes "Wireless technology has revolutionized access to local area networks when one can't always be close to an ethernet jack. But a recent research paper from the French Centre Nationale de la Recherche Scientifique tells how one slow user accessing a hot point can hurt the whole group. Apparently the very nature of CSMA/CD guarantees such anomalies. Here's the story, and here's the release from CNRS (in French)."
How long before this becomes an exploit in order to perform a DOS attack on wifi points?
I say about 1 month, maybe less. Any takers?
MMORPG fan-boy? Prove your worth
If you have an 11b AP with a bunch of PCs with 11 meg cards and some kind of 2 meg hand held device then the 2 meg device is going to hog the connection until its done.
A solution would be to leave the 802.11b AP inplace and servicing the older 2 meg devices and put an 11a or 11g AP in next to it to service the faster devices. Alternitively you might be able to put the slower devices on channel 1 and put the faster one on 6 and 11 (they have to be 5 apart to avoid overlap, and in the US you only have a total of 11 unregulated channels)
Wireless is different than wired communication. People are just going to ahve to get used to it. More stuff to learn, more rules to follow, more work for people like me.
Its a good thing- especially in this economy.
This isn't something that's likely to become a serious exploit, for the simple reason that the attacker would need to bring a compatible device within range of the access point. Unless someone has a serious grudge against the owner of the network, who'd want to spend the time? You'd either have to be present with a laptop/pda, or leave it behind. Remember, we are talking about radio waves here. Plenty of technology exists to track 802.11x signals, and all it takes is a well equipped sysadmin and a properly filed lawsuit to discourage the attacker.
Its good to see this sort of stuff put to paper anyway. As a Free/Open software implementer working on developing cheap-as-possible wireless access points for rural area internet distribution, I can say this IS useful. Problems like this, and many others, creep out of nowhere and are very hard to track down without expensive equipment. This specific problem happened and was remedied after much head-scratching by dividing one site into multiple cells so slow users had more 'time' to get their data. just my 0.02
jeff
Unfortunately, RTS/CTS is almost worthless in a situation that requires it - when you've got hidden nodes, the RTS mechanism can just as easily cause collisions.
The users of WAFreeNet (Perth, Australia) have just released some open source software (frottle) to combat this. Essentially it provides a polled/token operation at the IP layer, virtually eliminating collisions. This is a similar application to WiCCP, and we've been helping/competeing with the WiCCP developers. The other alternative is Karlnet Turbocell - expensive proprietarty software, firmware and hardware, with poor linux support.
I cant post any url's now - the websites wouldnt appreciate the slashdotting. For those of you than can find the sites for yourself, it may be well worth your time.
"People want a large range and high speed, it can't really happen"
Go back to your cave.
How much faster is wireless today then 5 years ago? A million percent? Do you think 802.11g is the end of wireless evolution? In ten years you don't think we will have gigabit wireless with 100m range?
Who needs sophisticated DOS attacks when simple jamming would do the trick quite well.
I'm a firm believer in the philosophy of a ruling class. Especially since I rule. -Randal, Clerks
Taking into consideration that Milwaukee is a city of ~600,000 people, that not many are going to have wireless laptops and even fewer are gonna spend time in a park surfing the web, it still seems like 20-40 users clogging a public WI-FI is a bit odd.
I suppose they can't complain, though, since the city paid less than a couple of hundred dollars for the setup. Still, it just seems somewhat pointless if it's gonna be clogged all the time.
"This food is problematic."
If I'm the only one on an 11mbps network and sitting right next to the station, I'd expect all the bandwidth. If somebody else joins the same network, in fairness my bandwidth may be halved. But do I really care whether the other guy is getting 1mbps or 11mbps during his timeslice? I'm still getting half the bandwidth as if I had it all to myself, right? I would only consider it strange if a single user joining with a weak connection cut my bandwith by > 1/n.
It sounds like it's time for some form of Token passing. That would prevent slow nodes from dominating the bandwidth, and would make sure everyone gets their fair share. Each node would have to register with the AP and then the AP would serve as the ring monitor.
--
Luck is just skill you didn't know you had.
All you need to shut down a wireless hotspot is a 2.4 GHz cordless phone. This will work for any of the 802.11X connections, lets give a hand to the FCC for this one.