Entering a word in a S-Box (like Feistel ciphers do) and keeping the S-Box secret: Could this be called encryption?

phantomcraft

8/10/23, 8:41 PM

Let's suppose I take a data block of 64-bits and use it as an entry in a (big) S-box, and keep the S-box secret. S-box would be the "key". For "decryption" I just enter the result of this operation in the S-box again to get the "plaintext"

For making all the blocks different each other in a sequence of them I would use a scheme like CBC, enter the first block in the S-box and XOR the result with next block and repeat this operation with next blocks. "Decryption" would be the inverse, enter the block in the S-box and XOR with preceding block.

Can this be called cryptography? Would this scheme deliver some security (neither if it is minimal)?

0 + 0

encryption

s-boxes

keys

kelalaka

8/10/23, 9:07 PM

AES froms as a family of permutations ( as all block ciphers) where each key selects one permutation ( not necessarily unique but expected). Now, once a key is set, we have a fixed permutation, can you see the relation between your design and this? Remember, we prefer simple designs to analyze...

phantomcraft

8/10/23, 9:42 PM

If my scheme delivers a decent security, it would be a very nice idea: Generate a random S-box with 16.384-bytes and use it as a key in which the 64-bits plaintext is entered, and use in CBC mode for making all the blocks different each other. This will (or would) deliver 131.072bits encryption. The only problem would be the very small block size, but I think there is methods to mitigate this.

LightBit

8/11/23, 11:25 AM

I probably did not understand, but it looks like a bad variant of Blowfish.

phantomcraft

8/13/23, 2:00 PM

@LightBit Yes, is bad variant of Blowfish. All that I need to know is if the data inserted as an entry in a S-box is recoverable is some way assuming the adversary doesn't know the S-box.

LightBit

8/13/23, 9:08 PM

It is susceptible to slide attacks, weak keys and related key attacks.

phantomcraft

8/13/23, 9:20 PM

@LightBit Ok, thank you for this reply! That's all I needed to know.

Elon Musk

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

EN: Entering a word in a S-Box (like Feistel ciphers do) and keeping the S-Box secret: Could this be called encryption?

TH: การป้อนคำใน S-Box (เช่น Feistel ciphers ทำ) และรักษาความลับของ S-Box: สิ่งนี้สามารถเรียกว่าการเข้ารหัสได้หรือไม่?

RO: Introducerea unui cuvânt într-o S-Box (cum o fac cifrurile Feistel) și păstrarea secretului S-Box: Ar putea fi numită criptare?

RU: Ввод слова в S-блок (как это делают шифры Фейстеля) и сохранение секрета S-блока: можно ли это назвать шифрованием?

VI: Nhập một từ vào S-Box (giống như cách làm của mật mã Feistel) và giữ bí mật S-Box: Điều này có thể được gọi là mã hóa không?

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.