-1

update!

if so, the NEW Q will be:

what is the strongest encryption algorithm ? (no time limits)
note that the algorithm can Not make more letters than the input

Community
  • 1
  • 1
D.christian
  • 167
  • 1
  • 9
  • 1
    Sounds like you want two things: 1) a good compression program (bzip2, rar, 7zip, xz, etc. come to mind) - most encryption algorithms don't do compression, and then 2) a good encryption algorithm. Note that many encryption algorithms use a fixed-length key generated from some form of user input - the user input can be any length, but it will get hashed/reduced/transformed in some way to a fixed length key (and possibly IV and other parameters as well). – twalberg Aug 27 '12 at 21:27
  • 1
    twalberg has the correct answer. Anything anyone else writes will just be repeating him - @twalberg, you should make that an answer. The only caveat is that I would also add [compression needs to be done before encryption](http://stackoverflow.com/questions/4676095/) – BlueRaja - Danny Pflughoeft Aug 27 '12 at 21:38
  • it seems that "super strong" should trump the other requirements. have you looked at AES 256 CBC ? http://www.appcove.com/passtool/ – gahooa Aug 27 '12 at 21:56
  • yes I did, but it will triple the amount of the text that will be sent.... I'm searching for a other solution, If I will not find one... AES will be the best for this case. – D.christian Aug 27 '12 at 22:01
  • I suggest you outline your actual requirements so that the experience of the collective SO brain can help you solve it. – gahooa Aug 27 '12 at 22:06
  • If your key is at least as long as your content, and you have a good pseudo-random key, then you could consider this: http://en.wikipedia.org/wiki/XOR_cipher Please read the caveats first, such as not re-using a key. – gahooa Aug 27 '12 at 22:08
  • Without clear requirements it is hard to say anything. E.g. it is unclear if a strong pseudorandom permutation would fit the requirements. On the other hand, one-time-pads are very easy to modify, very insecure against active attacks and hence not the right solution in almost all practical scenarios. – Jack Aug 28 '12 at 10:11

1 Answers1

2

http://en.wikipedia.org/wiki/One-time_pad

If your requirements are complete security and not increasing the length of the message at all, then it is possible to do this using a one-time-pad implementation.

Your key material needs to be:

  1. completely random
  2. the same length as your message
  3. never divulged
  4. never reused

If you are able to do these 4 things, then you can achieve maximum security without increasing your message length.

Your ability to implement this depends on your requirements. At the point you have done 1-4 above, you have not lessened the problem of key exchange (if indeed that is needed).

Read the linked article, and it will explain in good detail.

gahooa
  • 131,293
  • 12
  • 98
  • 101
  • the "completely random" could be a problem, but the "mathematically proven to be impossible to crack if used correctly" is exactly what I was looking for! thank you. – D.christian Aug 27 '12 at 22:35