Re: dwepcrack doesn't

Here's where I'm at today. Been running dwepdump since last sunday. I 
figure I'll let it run til next monday at least and see what happens. 
It's not looking good though. No answer from Cisco yet confirming 
whether or not the 340/350 series does weak iv filtering.

Periodically running dwepcrack -w -f <various>.

Interesting thing about brute force. I was able to brute force the WEP 
key I'm using (0101010101) after about 6 hours of packet capture. Since 
then, I have periodically run dwepcrack -b logfile or -b -w and it 
hasn't been able to crack it again. A very short packet capture didn't 
work and longer ones have failed too. I would have expected a brute 
force to run until it guessed the key or I got tired of watching it. 
Could be dwepcrack picked a lucky packet to crack the first time.

Tye

============
2-27-03

logfile is up to 460 mb

# ./dwepdump  -w wi0 logfile

* dwepdump v0.2 by h1kari <[email protected]> *
* Copyright (c) Dachb0den Labs 2002 [http://dachb0den.com] *

starting pcap capture loop...
  device: wi0
  logfile: logfile


0:112 1:10ca2 2:1a391 :: 8505274



# ./dwepcrack  -w -f 4 logfile

* dwepcrack v0.4 by h1kari <[email protected]> *
* Copyright (c) Dachb0den Labs 2002 [http://dachb0den.com] *

reading in captured ivs, snap headers, and samples... done
total packets: 463114

calculating ksa probabilities...
  0: 8/768 keys (!)
  1: 2048/131328 keys (!)
  2: 3076/197376 keys (!)
  3: 3076/197120 keys (!)
  4: 5379/328703 keys (!)

(!) insufficient ivs, must have > 60 for each key (!)
(!) probability of success for each key with (!) < 0.5 (!)

warming up the grinder...
  packet length: 40
  init vector: 01:03:04
  default tx key: 0

progress: ..................................



h1kari wrote:
> Tye,
> 
> It looks like the access point is doing some sort of weak iv filtering
> because you're getting a hella lot of duplicates. The numbers with dwepdump
> is the total number of packets with weak ivs for the different key bytes,
> whereas the numbers in dwepcrack are the number of unique weak ivs (which as
> you see are significantly smaller). It might still be possible to crack it,
> but you're going to need to collect a lot more packets to get enough unique
> weak ivs.. I'll try and build some fast weak iv checking into dwepdump and
> such in the next release so it's a little more intuitive.
> 
> I've been able to crack hardware that isn't doing weak iv filtering fairly
> fast in both lab and live environments. The iv filtering stuff is tricky,
> since I don't know exactly what sort of algorithm the APs and cards are
> using to avoid weak ivs. Either way it's really difficult to filter all of
> the weak ivs and still have decent network speed and still avoid iv
> collisions, so I'm sure there's still some that are sent out.
> 
> I'd be interested to see if you're able to crack it by capturing more
> packets. Please let us know if you can get it to work..
> 
> Thanks,
> -h1kari
> 
> 
> On 2/24/03 17:33, "Tye F. Hammerle" <[email protected]> wrote:
> 
> 
>>I'm working on cracking wep keys with OpenBSD 3.2 using the bsd-airtools
>>package.  I've gathered more than enough packets according to what I've
>>read yet it seems uncrackable.
>>
>>I've got a Cisco 340 Access point setup with 40-bit wep (key =
>>0101010101 ) no tkip or mic enabled. I have another PC setup with a
>>Cisco 350 wireless card. The PC uses a batch file to copy and delete a
>>65mb file to a server across the wireless network. It's been running for
>>a couple of days now and here's what I've got.;
>>
>>
>>#./dwepdump  -w wi0 logfile
>>
>>* dwepdump v0.2 by h1kari <[email protected]> *
>>* Copyright (c) Dachb0den Labs 2002 [http://dachb0den.com] *
>>
>>starting pcap capture loop...
>>device: wi0
>>logfile: logfile
>>
>>0:6c 1:6589 2:9e5d :: 3183449
>>
>>
>>According to the above I've got 6c (102 dec) weak packets, well over the
>>40 (60 dec) suggested as necessary in the docs.
>>
>>Now lets see if we can crack it;
>>
>>#./dwepcrack  -w   logfile
>>
>>* dwepcrack v0.4 by h1kari <[email protected]> *
>>* Copyright (c) Dachb0den Labs 2002 [http://dachb0den.com] *
>>
>>reading in captured ivs, snap headers, and samples... done
>>total packets: 174324
>>
>>calculating ksa probabilities...
>>0: 8/768 keys (!)
>>1: 2048/131328 keys (!)
>>2: 3076/197376 keys (!)
>>3: 3076/197120 keys (!)
>>4: 5379/328703 keys (!)
>>
>>(!) insufficient ivs, must have > 60 for each key (!)
>>(!) probability of success for each key with (!) < 0.5 (!)
>>
>>warming up the grinder...
>>packet length: 40
>>init vector: 01:03:04
>>default tx key: 0
>>
>>progress: ................................................
>>
>>
>>Yet when I run Dwpecrack, it gets nowhere. I've tried -f 1 through 9. I
>>also don't understand the "must have > 60 for each key" what is the n:x
>>counter (x) for? I thought that was the weak key counter and I've got 6c
>>which is way over the 60.  I figure I'm confused.
>>
>>Before I got this output I ran another capture for about a day getting
>>over 20 on (x). Didn't crack. I ran dwepcrack -b and brute forced the
>>key in 16 hours. (my laptop is a PII-333)
>>
>>So, I'm looking to find out what I should be expecting to see for
>>numbers from dwepdump. Also,  how much of a given is it that once you
>>have enough weak keys you can crack the key?  Based on what the docs
>>tell me and the little I think I have learned, I am getting enough that
>>dwepcrack should have cracked the key.
>>
>>I've read several lists where the people there are saying that there is
>>a significant difference between cracking a real live network verses a
>>lab test network. Supposedly the real network cracks alot easier.
>>Interesting.
>>
>>Where am I going wrong?
> 
> 
> _______________________________________________
> Bat mailing list
> [email protected]
> http://lists.dachb0den.com/mailman/listinfo/bat