So wanted to check in with you guys to see if anyone can point out any underlying faults in a SPNA cipher strcuture.
Base idea is simple. SPN based like AES-256 however algebraic. No lookup tables. No state machines. No branching. No memory. Only AND, NOT and XOR logic and butterfly mixing.
Has anyone tried this before? I have made one and have it published on my github which fully works but i want the idea sanity checked.
If your wundering its a 64 byte block cipher in CTR mode.