(The following situation is hypothetical.)

I need to communicate something securely and I have decided to use one-time pads.

I am going to be using human genes as the source of the OTP data (this is important).

If I were to use (for example) a weak kind of encryption, in this case a variation of XOR 'encryption', would the fact I am using OTPs offset the weakness of XOR 'encryption' or would it be a complete waste of time because XOR is still inherantly weak?

If it matters, before being XOR 'encrypted' the messages will be plaintext in an unspecified human language (English, Japanese et cetera - the language is irrelevant other than being a known, well spoken human language).

Disclaimer: I won't be offended if I'm told that this is a horrible idea - It's not my idea.

  • 3
    $\begingroup$ What you are describing really isn't OTP because genes don't satisfy the requirements of a key for use in OTP. $\endgroup$ – kasperd Sep 19 '17 at 21:23
  • 2
    $\begingroup$ Also genes as a source is a bad idea as the reading process may not always give the exact same data. Its a lot of mixing bowls and jelly and cleaning, way different to plugging in a disk that has error correction. This means if you intended to use a humans DNA at end A as the key to encrypt, and then use the same human after re-sampling at B you probably don't have the same key. $\endgroup$ – daniel Sep 20 '17 at 11:50

OTP requires a completely random key. That's next to impossible, and human genes are certainly not completely random. This is why OTP's are mainly theoretical constructs.

XOR is the normal way of encrypting using an OTP. XOR is not weak encryption if the key stream is strong. And if the key stream is perfectly secure then XOR encryption is completely secure (for confidentiality anyway).

And no, the messages do not matter. A secure cipher should be secure against known plaintext attacks. That is, different messages or other parts of the message should remain confidential even if part of the message is known.

It's good that you're not offended, because it certainly is a horrible idea.

  • 3
    $\begingroup$ Nitpick: the difficulty of distributing a key the same size as the plaintext securely is a much harder problem than having a good source of random numbers. Having good RNGs isn't a strong contributor to not using OTPs. I know that you know this, you've said as much in the past IIRC. $\endgroup$ – SAI Peregrinus Sep 19 '17 at 20:37
  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$ – e-sushi Sep 19 '17 at 23:13
  • $\begingroup$ But what is this: researchgate.net/figure/… ? It seems to be a genetic encryption system. I'm unable to read any further details though... $\endgroup$ – Paul Uszak Sep 21 '17 at 22:15

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.