DESX Whitening
From: Sebastian Faust (sfaustdeleteme@logic-software.de)
Date: 02/19/03
- Next message: Lew Pitcher: "Re: Encrypting again an already encrypted file increase security ?"
- Previous message: Marvin Cohen: "Re: Encryption of a password that is used to encrypt code - a bad idea?"
- Next in thread: David Wagner: "Re: DESX Whitening"
- Reply: David Wagner: "Re: DESX Whitening"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
From: "Sebastian Faust" <sfaustdeleteme@logic-software.de> Date: Wed, 19 Feb 2003 21:39:28 +0100
Hi,
I read in the book from Bruce Schneider "Applied Cryptography" that when
whitening is used in the DESX case a brute-force attack requires 2^(n+m/p)
operations if K_1 = K_3 or 2^(n+m+1) if K_1 and K_2 are different and we
have 3 known plaintexts, where n is the key size, m the b lock size and p
the number of known plaintexts.
DESX is in that book defined as followed:
C = K_3 + E_(K_2)(P + K_1),
P = K_1 + D_(K_2)(C + K_3),
where E_(K_2) is encryption with DES using K_2 and D_(K_2) is decryption.
Now my question: Why 2^(n+m/p) and 2^(n+m+1)? And what has it to do with the
number of known plaintexts?
Thanks in advance
Sebastian
- Next message: Lew Pitcher: "Re: Encrypting again an already encrypted file increase security ?"
- Previous message: Marvin Cohen: "Re: Encryption of a password that is used to encrypt code - a bad idea?"
- Next in thread: David Wagner: "Re: DESX Whitening"
- Reply: David Wagner: "Re: DESX Whitening"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
Loading