1
$\begingroup$

This question already has an answer here:

Are there any algorithms that exist to encode messages written with pen and paper such that the messages can be encrypted/decrypted mentally? For example, Bob thinks of the text "This is my plaintext", writes down "siht si ym txetnialp", and then Larry (who knows the cipher) is able to understand the message "This is my plaintext", just by looking at the cryptotext. The algorithm is simple enough that nothing needs to be written down besides the cryptotext.

This should not be a code, where a specific phrase means something else. I'm looking for an algorithm like a one-time-pad or modular addition method that can be completely contained in one's head.

The strength of the encryption is not as important - it doesn't need to stand up to a complex solver, just be very difficult to decode by someone who doesn't know the cipher. The more important feature here is speed - e.g. encrypting and writing a message can be done simultaneously in (on the order of) twice the time it takes to write it in plain English.

$\endgroup$

marked as duplicate by e-sushi Jul 1 '15 at 0:45

This question has been asked before and already has an answer. If those answers do not fully address your question, please ask a new question.

  • $\begingroup$ Similar to this questin. And the answer is that you can do RC4 if you're willing to write a few things down, Solitaire and its variants if you have a deck or cards, or a one time pad if you really have nothing else. Why have the strict requirement of it taking at most twice the time of writing normally? That's 22 wpm, according to wikipedia, which you can't hope to come even within an order of magnitude of. $\endgroup$ – robertkin Jun 30 '15 at 18:00
  • $\begingroup$ Any mono-alphabetic substitution cipher can probably be done mentally, but it would not be much different from substituting words and hence would be easy to break (even mentally). But in case you think it does suffice (i.e. what you want is obfuscation), just using alternate symbols to write the message down (and maybe making changes to the spacing) is an option. $\endgroup$ – Aleph Jun 30 '15 at 18:04
  • $\begingroup$ @Aleph - great word, obfuscation; I had to look it up. I suppose I was being overly optimistic about manual encryption speed, so I'll have to make due with mono-alphabetic substitution. Thanks for the ideas. $\endgroup$ – lg22woo Jun 30 '15 at 19:15
  • $\begingroup$ speck (en.wikipedia.org/wiki/Speck_(cipher)#Reference_code) in counter mode is easier and more secure than RC4... but you still need to use some sort of MAC, no matter what cipher you use. $\endgroup$ – lily wilson Jun 30 '15 at 21:46
  • $\begingroup$ Welcome to Crypto.SE. I´m sorry to say so, but as it seems the same question has been asked before… which is why I closed this one accordingly. In case you want to edit your question to clarify the difference between your question and the other one, please feel invited to do so. When done, you can request your question to be reopened with a simple click on the “reopen” link. In all other cases, please feel invited to ask a new question. $\endgroup$ – e-sushi Jul 1 '15 at 0:55