Let’s say there’s a language that has all the letters of the alphabet evenly distributed.

Is this language then secure against frequency-analysis or could there still be some soft of weakness, i.e. by using too short keys?

  • $\begingroup$ What is the encryption scheme? $\endgroup$ – Shan Chen Sep 8 '18 at 20:41
  • $\begingroup$ subsituition (ceaser / vingenère) $\endgroup$ – AleksanderRas Sep 8 '18 at 20:49

Substitution cipher (e.g., Caesar) and polyalphabetic substitution cipher (e.g., Vigenere) are deterministic encryption schemes. The same plaintext will always result in the same ciphertext. So, they are not secure in the sense of indistinguishability under chosen plaintext attack (IND-CPA).

It may be possible that "all the letters of the alphabet are evenly distributed", but it is very unlikely that all the 2-grams, 3-grams, etc., are evenly distributed (because some of them may be meaningless in the language). In this case, the frequency analysis can still work.

| improve this answer | |
  • 1
    $\begingroup$ What’s 2-grams, etc.? Sadly Google only returned the price for weed when I tried to search this (cries cryptographically) $\endgroup$ – AleksanderRas Sep 8 '18 at 21:10
  • 1
    $\begingroup$ @AleksanderRassasse In general, an n-gram is a contiguous sequence of n letters. For instance, here are some 2-grams in English: is, as, he, my. $\endgroup$ – Shan Chen Sep 8 '18 at 21:17

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.