Autokey

I was wondering if the variant of the Vigenere Autokey cipher I will
shortly describe is effective at strengthening the cipher, or if it
actually weakens it. My gut reaction is to assume the latter, but it
seems as if it may work to distort the periodicity of the key, which is
(by my understanding) the prime method of attack on such ciphers. So I
come here to inquire.

The key is a string which is split up into two* or more numbered
sections, beginning at 0.
*(If there is only one section, this cipher is equivalent to the
(non-autokey) Vigenere cipher.)

The intial key section is assumed to be the first (section 0), and it
is applied to the plaintext as in the Vigenere cipher until it is
exhausted (not repeating it). The last letter of the generated
ciphertext is then taken (modulo the number of sections in the key) to
indicate the next section of the key. Repeat until plaintext is
exhausted.

Decryption is trivially the reverse of this process.

I hope that was clear enough and I will appreciate responses!

-- Ben

[Base64 Autokey]
Received on Fri Dec 23 20:09:56 2005