Prev: Announcing: First-ever on-line collection of declassified Chaocipher-related correspondences
Next: Announcing: First-ever collection of declassified Chaocipher-relatedcorrespondences
From: Mok-Kong Shen on 2 Oct 2009 09:48 Hi, A recent post of mosherubin on the distinguished Chaocipher reminds me of the methodologies of classical crypto in general and as a consequence also of a conversation I had long time ago with a friend. We babbled one day over the application of polyalphabetic substituion ciphers and had as a topic in our chat the following variation of the use of substitution tables, which, though we were convinced must with very high probability had already been tried out in theory or in practice (and possibly even subsequently discarded due to disadvantages), we hadn't known of a mention in the meagre collection of books then available to us. Let there be chosen two 26x26 substitution tables T1 and T2, whose columns are random permutations of the alphabet. The encipherment with table T of the plaintext character P with the key character K will be denoted by C = T( K, P ), resulting in the ciphertext character C. The encryption of the plaintext character sequence P_i (i=1 ...) is then, with K_1 being supplied at the start: K_i = T1( C_(i-1), P_(i-1) ) C_i = T2( K_i, P_i ) A special case is where T1 and T2 are identical. My friend and I hadn't pondered too much over the scheme (our coffee cups were soon empty), let alone actually tried it out. I myself have almost forgotten it and have only slowly retrieved it from the 2nd order storage of my poor brain. I am taking the liberty to post it here, because I suppose it could eventually serve as a stuff for recreation for those experts interested in classical cryptology. Thanks, M. K. Shen --------------------------------------------------------------------- Was sich ueberhaupt sagen laesst, laesst sich klar sagen; und wovon man nicht sprechen kann, darueber muss man schweigen. L. Wittgenstein |