Timeline for Advantage of $\operatorname{GF}(2^8)$ over $\mathbb Z/2^8\mathbb Z$ in AES/Rijndael
Current License: CC BY-SA 3.0
4 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jan 7, 2017 at 22:09 | comment | added | CurveEnthusiast | Well the question seems to infer that working in $GF(2^8)$ is unique to MixColumns, while it is used through all of AES. If you replace all $GF(2^8)$ operations by $\Bbb Z/2^8\Bbb Z$ operations, what you end up with does not make sense. Yes, you can forget about this and only look at MixColumns, and then yes, perhaps you could use an invertible matrix. But is it really worth doing that without having the big picture in mind? | |
| Jan 7, 2017 at 20:32 | comment | added | poncho | I don't see how this answers the question; yes, there are noninvertable (nonzero) elements in $Z/256$, however that doesn't prevent nontrivial invertable 4x4 matrices. Yes, you couldn't use $Z/256$ in the ByteSub operation; however the question wasn't about the ByteSub operation. | |
| Jan 7, 2017 at 16:43 | history | edited | CurveEnthusiast | CC BY-SA 3.0 |
added 9 characters in body
|
| Jan 7, 2017 at 16:36 | history | answered | CurveEnthusiast | CC BY-SA 3.0 |