Weak IV's

[mfsu21]

I'm trying to analyze airsnort and how it works. I understand it uses the FMS theory to get the WEP key but what I don't understand and can't seem to find a definition that I understand, what is considered a weak IV? I have the definition in the documentation of AirSnort.

"Useful packets defined by Schmoo are those packets where the first byte in the Initialization Vector is a number three greater than one of the offsets of the bytes of the key. For 128 bit encryption, this means a number from 3-16. The second byte must be 255 and the third byte can have any value. This means that for every byte of the key, there are 256 weak IVs."

Now I've also read that wireless cards based on the Lucent technology reset the IV everytime the card is removed and inserted into the laptop. Are the lower number IV's weak IV's just because there is a possibilty of them reocurring more often? Or is there some senerio where the key is just weak like in the Schmoo definition?

Thanks for the input.
Mike

[mfsu21]

I found a paper explaining weak keys in RC4. It looks like this is what AirSnort uses to find the weak keys.

http://marcel.wanda.ch/Archive/

Download Andrew Roos paper.

[BER_vs_SNR]

The IV being of 24 bits length only, one can see its value space easily exhausted by a moderately active AP providing one with ample munition for an attack. Look at (Berkeley's) Nikita Borisov's slides for rough time estimates of how easy it is:

http://www.isaac.cs.berkeley.edu/isaac/wep-faq.html

Also their paper presented at Mobicom shows you some of these attacks in detail;

http://www.isaac.cs.berkeley.edu/isaac/mobicom.pdf

You may also want to understand how the set up of RC4 makes the key scheduler compromisable as you concatenate IVs with basic key to come up with the actual encryption key. A pioneering work by Scott Fluhrer, Itsik Mantin, and Adi Shamir identified a flaw in the RC4 key setup algorithm which results in a total recovery of the secret key. Implementing the attack requires the collection of traffic passively. Their paper is at:

http://www.cs.umd.edu/~waa/class-pubs/rc4_ksaproc.ps