Re: Encrypting twice with the same key

From: lurker (na@nospam.org)
Date: 04/13/03


From: na@nospam.org (lurker)
Date: Sun, 13 Apr 2003 14:44:45 GMT

On Sun, 13 Apr 2003 15:48:53 +0200, "Daniel Weber"
<ComeAndSpamIt@gmx.net> wrote:

>Is encrypting twice with the same key safe?
>I made up the following:
>
>May f be the encryption transformation, and A be its Matrix with entries
>from Z.
>f(x) -> Ax
>
>Then encrypting twice would be f(f(x)) = (AA)x = (A^2)x.
>
>Now, there's much lesser possibilities for the entries of A^2.
>That would make brute force a better choice.
>
>Did I think right or am I just insane?
>
>Daniel
>
>
In a good encryption algorithm a key used twice would reduce the
number of possibilities by a factor of 2.

In the overused example of the classic one time pad if the plaintext
is known or can be guessed then the key is known.

To estimate the theoretical strength of the cipher you are using you
should see how often the session key is replaced with a new random
key.