0

Today, in class we were discussing about how s-box is used in feistal ciphers. Our sir went on to say that for s-box substitution, we take the 1st and last bits as one part and the remaining bits as another, following which we find the matching 4 bits.

enter image description here

So if we were to have the 6bits as: "100100", it would map to 1110 (left column: 10 , right column: 0010)

My doubt now is if it possible for us to take different positions of bits together, like the 1st and 3rd bit? If not, is there a reason why we can only take the 1st and last bits?

jeremy
  • 59
  • 7
  • 2
    It was the designer's choice that they choose this analyzed the design. You can interpret whatever you can. however, the algorithm will differ and it will need a new analysis. Normally you need to ask this in [cryptography.se]. This is basic and you got your answer here. – kelalaka Mar 01 '22 at 13:57
  • Thanks for the answer and ohh damn, my bad. Will keep this in mind when i have some similar doubt. But also, i have a follow up doubt, how would you analyze a design though if it's better or good? Is that simply based on no of tries you need to break a cipher? – jeremy Mar 01 '22 at 16:16
  • 2
    Well, that is a long history that NSA was aware of the Differential attack. Researchers conclude this after noticing that the changes of the NSA make the cipher better. Researchers also believe that NSA was not aware of the Linear Attack of Matsui. Your actual question has deep knowledge about attacks on block ciphers. You might look at the books like block cipher companion and Desing of Rijndael. ( search for them). – kelalaka Mar 01 '22 at 16:39
  • woww, a lot of stuff idk . But its all interesting. Gonna be looking up all the stuff you mentioned. Thanks a ton for getting back! – jeremy Mar 01 '22 at 17:24

0 Answers0