Re: Weakness of Feistel ciphers



In article <06kih3-94n1.ln1@xxxxxxxxxxxxxxxxxxxxx>,
Kristian Gjøsteen <kristiag+news@xxxxxxxxxxxx> wrote:
Kim G. S. Øyhus <kim@xxxxxxxxxxx> wrote:
And how am I to know how many times i must iterate to decrypt?

If you think for a moment, you realise that you should stop when
you get a result less than the modulus. That is how many times
you must iterate.

And that is precisely what I thought tonight, in bed, away from Usenet.

Nice method, but with 2 disadvantages I can see now:

1. It is definitely slower than the cipher it is built upon.

2. It is only suitable for small decrements in the logarithm of its range.
i.e. it works badly if one wants to do 128bit -> 100bit because it
gets extremely slow.

I might still use the method some time, f.ex. in monte carlo integration,
but not now.

Kim0
.