# Determining Involutary Keys in Substitution Ciphers

How would I be able to show that a key $K=(a,b)$ is involutary for an Affine Cipher over the integers modulo n?

Similarly, is there a way to generalize the conditions for a key to be involutary for a substitution cipher?

(Involutary keys are keys for which the encryption function is equal to the decryption function $e_K=d_K$)

• Comments are not for extended discussion; this conversation has been moved to chat. – e-sushi Feb 2 '16 at 22:15

If $E(x) = ax+b \bmod n$, and $E(E(x)) = x$, then you need $a(ax+b)+b = x$, or $a^2x + ab + b = x$. So any (a, b) where $a^2 = 1$ and $b(a+1) = 0$ will do it. There are always solutions $(1, 0)$ (which corresponds to not encrypting) and $(-1, b)$. Depending on the value of $n$ there may be other solutions for $a$. There are only those two for $n = 26$, but there are a total of 4 solutions for $n = 36$.
• To find the solutions, are you just testing that each element of (say) $\mathbb{Z}_{15}$, satisfies the above conditions or is there another method being used? – Cryptkeeper Feb 3 '16 at 5:24
• For small values of n, that's easiest. For larger, I think you can use the factorization of n. – bmm6o Feb 3 '16 at 13:40
Any substitution cipher $E$ over a set $M$ might be expressed as a permutation $\pi_E$ of the elements of $M$. Any permutation is uniquely identified by its decomposition into cycles, save for the order of the cycles. The condition $E(E(x)) = x$ will be met for all $x \in M$ if and only if the permutation $\pi_E$ is decomposed into cycles of length 1 or 2 only.
• For instance, if I had some sort of system where I was working over $\mathbb{Z}_4$, the conditions of the substitution cipher is $S_4$? (where $S_4$ denotes all possible permutations of elements in the symmetric group with four elements)? – Cryptkeeper Feb 3 '16 at 7:31
• Sorry, I should've asked, is it all elements of $S_4$ with order $2$? – Cryptkeeper Feb 3 '16 at 8:19
• The symmetric group $S_n$ is defined as the set of all permutations over the elements of $\mathbb{Z}_n$ with composition of permutations being the group operation. If and only if an element in such a group is of order 2, then it meets the condition in your question, yes. – Henrick Hellström Feb 3 '16 at 10:00