ε-close to t-wise inependence of SPN (AES)

hardyrama

11/22/22, 10:55 AM

According to theorem 3.13, the 6 rounds of AES is 0.472-close to pairwise independence. It is also mentioned t-wise independence used to analyze higher order derivates attacks. it is also mentioned 3-wise indepdent permutations have a potential application in strengthening short encryption keys. My questions are related to the fundamentals of t-wise independence permutations.

Q.1 What does it mean to have 2, 3 or 4-wise independence to a block cipher from a mathematical perspective?

Q.2 what does it mean 6 rounds of AES is 0.472-close to pairwise independence?

Q.3 (assumption) Does higher order independence have opposite relation with key size but same security level?

0 + 3

cryptanalysis

aes

spn

Mikero

11/22/22, 2:41 PM

Q2: Definition 2.5 explains what it means to be $\delta$-close to t-wise independence. Q1: t-wise independence is Definition 2.5 with $\delta=0$

kodlu

11/23/22, 2:23 AM

your questions are a bit vague and some answers are embedded in the definitions as in the comment by @Mikero. Also please help readability by stating at least one of these definitions.

hardyrama

11/27/22, 2:47 PM

Thanks for pointing to the definition but what implication does it have in term of AES security (number of rounds), secondly, for Q1, is there a relationship between k-wise and higher order derivatives (for example 2-wise means immune against second order differential cryptanlysis)

Elon Musk

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

EN: ε-close to t-wise inependence of SPN (AES)

TH: Îµ-ใกล้เคียงกับความไม่เป็นอิสระของ SPN (AES)

RO: μ-aproape de independența t-wise a SPN (AES)

RU: Îµ-близкая к t-независимость SPN (AES)

VI: Îµ-gần với sự phụ thuộc t-khôn ngoan của SPN (AES)

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.