how to build textset for Impossible Differential Cryptanalysis on IDEA?

siba36

12/14/23, 9:18 AM

I'm implementing impossible differential cryptanalysis on 3.5 round IDEA using the methods in "Miss in the middle attack on IDEA and Khufu" paper.

In the first step, I need to provide $2^{32}$ pairs that are identical in $X_2$ and $X_4$ and have all the possibilities in $X_1$ and $X_3$ to start the cryptanalysis.

So I've tried to collect the text for various books to provide the first plaintext and the second plaintext of the pair is computed so that it matches the input difference of the pairs.

But all the text that I come up with is equivalent to only 2 million plaintexts so I'm repeating this for the rest of the pairs to reach $2^{32}$ pairs.

Is this correct or should I have to get unique text for each pair and if that's what I should do where I could find these large texts?

1 + 0

differential-analysis

idea

Score:1

Crypto

kodlu

12/14/23, 1:24 PM

Modern cryptography is based on security analysis assuming no weaknesses in plaintext, i.e., no characteristics of English or any other natural language.

So you don't need to use "books" (if I understand what you are doing correctly) to generate plaintexts with the difference structure required. You generate plaintexts at random with the prescribed structure.

So you just iterate through structures pseudorandomly.

Iterate $2^{32}$ times for $i=1,\ldots,2^{32}$ and for each structure (list) $S_i=[]:$
Pick $X_2$ at random (and set $X_4$ equal to it)
Loop: Pick $X_1$ and $X_3$ over all possibilities
Append $[X_1,X_2,X_3,X_3]$ to the structure $S_i$
End Loop
Your structure $S_i$ is ready
End

+ 0

Elon Musk

I sit in a Tesla and translated this thread with Ai:

EN: how to build textset for Impossible Differential Cryptanalysis on IDEA?

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.