The attack is not related to an reused IV. This attack works fine if 40,000 packets have been collected, even if not a single one uses an IV which had been used previously. Some 802.11 stacks use a monotonic counter for IV generation, which will never repead an IV in less than 2^24 packets. Even if such a counter is used, this attack will work perfectly.
Sorry, but this is really a new attack which doesn't need a million IVs anymore for 128 bit wep. Instead a number of 40,000 to 60,000 is most times sufficient to calculate the secret key.
It is not an april fool. It is correct that the paper was Submitted on April 1., and later updated on April 3..
If you still think this is an april fool, you can download and run the tool on some captured data, and see how it works.
Hi
This is attack is not based on the original WEP attack anymore and doesn't know the concept of weak IVs. I think WEPPlus won't have any noticeable impact on the successrate of the attack.
If you managed to load the paper, have a look at the plot of the successrate of the paper. The graph labeld linux iv keylimit 1000000 should be more or less exactly the successrate the attack should have in an WEPPlus environment. This rate doesn't differ much from the optimal rate.
It is a little bit WEPs fault. I think RC4 can still securely be used. For examle in TKIP or in SSL/TLS the RC4 key generation differs from the algorithm WEP uses, and can still be seen as secure.
Slashdot Mirror
The attack is not related to an reused IV. This attack works fine if 40,000 packets have been collected, even if not a single one uses an IV which had been used previously. Some 802.11 stacks use a monotonic counter for IV generation, which will never repead an IV in less than 2^24 packets. Even if such a counter is used, this attack will work perfectly.
Sorry, but this is really a new attack which doesn't need a million IVs anymore for 128 bit wep. Instead a number of 40,000 to 60,000 is most times sufficient to calculate the secret key.
It is not an april fool. It is correct that the paper was Submitted on April 1., and later updated on April 3.. If you still think this is an april fool, you can download and run the tool on some captured data, and see how it works.
Hi This is attack is not based on the original WEP attack anymore and doesn't know the concept of weak IVs. I think WEPPlus won't have any noticeable impact on the successrate of the attack. If you managed to load the paper, have a look at the plot of the successrate of the paper. The graph labeld linux iv keylimit 1000000 should be more or less exactly the successrate the attack should have in an WEPPlus environment. This rate doesn't differ much from the optimal rate.
It is a little bit WEPs fault. I think RC4 can still securely be used. For examle in TKIP or in SSL/TLS the RC4 key generation differs from the algorithm WEP uses, and can still be seen as secure.