Score:1

Chaos-based AES, is it secure?

us flag

In the link below, the author uses the aes as a basis for his cipher. In his words: The thesis investigates and explores the behavior of the AES algorithm by replacing two of its original modules, namely the S-Box and the Key Schedule, with two other chaos-based modules.

One might ask: will this system at least will inherit the security I the aes? In addition, it is a common theme in chaos-based ciphers take the outlook of the conventional cipher, so what is fundamentally wrong with chaos-based ciphers?

https://repository.kaust.edu.sa/bitstream/handle/10754/292821/Naif_Thesis.pdf

fgrieu avatar
ng flag
One kind of "proof" proposed in the two linked thesis/articles is that the output (for consecutive inputs, I guess) pass NIST's statistical test. That's an abysmally wrong (yet very common) security argument, and goes a long way to discredit the whole work.
user2357 avatar
us flag
@fgrieu I the this is the case for almost all chaos-based systems. However, the article I put in the comments says that the scheme is provable (not just statistical). Is this is the case?
fgrieu avatar
ng flag
In that [second paper](https://www.mdpi.com/2073-8994/13/5/824) I see _claims_ of mathematically provable security. But no statement of a proposition to prove, or proof. Do you? Make your own opinion about if this paper makes unproven claims.
user2357 avatar
us flag
@fgrieu I see your point. Thank you.
Paul Uszak avatar
cn flag
I understand your reservations, and I know the consensus on this forum. Yet. Chaos = entropy, which leads to cryptography. S-boxes don't have to be that clever as security can be strengthened with additional rounds. Deterministic chaos is a proven and well studied field, both mathematically and electro-mechanically. I believe the dissertation has merit.
Maarten Bodewes avatar
in flag
Chaos is not the same thing as entropy, especially not when it is, in the end, created by deterministic algorithms. Please note that Paul's definition of Chaos and Entropy are at odds with most of the people in this community. That doesn't disqualify the paper in any way and I agree that the paper in the question may well have some merit.
mangohost

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.