davidmichaels wrote:Can any one explain this to me :p
Very briefly, an encryption system involves taking a message you want to send and a 'key', which is some extra information which doesn't convey any message, and combining them in some way to form an encrypted message that is what you actually send. As a very simple example, the key might be an English word or phrase with no repeated letters, such as "THE QUICK BOARS", and the encryption method might be to write down the alphabet, then below it the phrase followed by all the letters of the alphabet that are not in the phrase in reverse alphabetical order:
A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
T H E Q U I C K B O A R S Z Y X W V P N M L J G F D
Then take the message you want to send (which had better be written in a form that makes sense with all its spaces and other non-letters omitted, and doesn't care about the difference between capital and lower-case letters) and change each letter to the letter that is below where it appears in the top row. For example, suppose the message you want to send is "The enemy will land at 1:30 pm tomorrow". You'd better change the numerals to words and check that it's still readable when written without spaces or the colon, and all in capitals "THEENEMYWILLLANDATONETHIRTYPMTOMORROW" (it's not ideal, but the recipient should be able to work out e.g. that "ATONE" is supposed to be "at one" and not "a tone" because one makes a sensible message and the other doesn't!). Then translate it according to the table to produce "NKUUZUSFJBRRRTZQTNYZUNKBVNFXSNYSYVVYJ".
As long as the recipient knows the key "THE QUICK BOARS", they can easily write down the same table and decrypt the encrypted message by changing each letter to the letter that is above where it appears in the bottom row. Anyone else has a harder job...
The problem is making certain the recipient knows the key! Sometimes that's easy, but this problem supposes they don't and have got to be told it. Rover110's solution involves producing four different keys, encrypting the message first with one key and then encrypting the encrypted message again using a different key, in four different ways, and giving each doubly-encrypted message to one of the messengers along with one of the keys that wasn't used for either of its encryptions. Furthermore, the exact scheme is chosen in such a way that the enemy, who gets two of the doubly-encrypted messages and two of the keys, never gets the right two to decrypt either of their doubly-encrypted messages, but the recipient, who gets three of the doubly-encrypted messages and three of the keys, always gets the right two to decrypt at least one (and in fact always exactly one) of their doubly-encrypted messages.
I should probably add that I've selected the example encryption technique I've used above to be easily explained,
not to be a good encryption technique! It most definitely isn't a good encryption technique: just the knowledge that it's a simple letter-for-letter substitution and that the encrypted message is "NKUUZUSFJBRRRTZQTNYZUNKBVNFXSNYSYVVYJ" (i.e. without knowledge of the key) is enough to allow a suitably-programmed computer with a comprehensive list on English words to decrypt it by systematically checking the possibilities: the first word must be a one-letter word, or a two-letter word made up of two different letters, or a three-letter word made up of three different letters, or a four-letter word that starts with three different letters and then repeats the third of them, etc. For each of those possibilities, it can run through its dictionary looking for words that match; for each such match, it can then work similarly with a more restricted set of possibilities for the second word (for example, on what is in fact the correct for word "THE", the possibilities for the second word are the one-letter (non-)word "E", the two-letter words consisting of "E" followed by a letter that isn't in "THE", the three-letter words starting and ending with "E" and with a letter that isn't in "THE" in the middle, etc. The list of possible first words is quite long (probably tens of thousands, I would guess), but the list of possible second words for any particular one of them will generally be a lot shorter, the list of possible third words shorter still, etc, and so on: overall, the number of possibilities it needs to explore is probably only in the millions to billions range, which means such a computer can come up with a short list of possible messages in a matter of seconds to minutes.
As an example, try feeding "NKUUZUSFJBRRRTZQTNYZUNKBVNFXSNYSYVVYJ" to
https://quipqiup.com/, which is a publicly available solver for such simple letter-for-letter encryptions (I don't think it works in quite the way I describe above, by the way, but there are plenty of methods for breaking such encryption techniques). It doesn't get the message quite right, guessing that there's a space where there isn't, but the enemy would have no trouble seeing what it was intended to be and deciding to change the time of their landings!
But modern encryption techniques are much more sophisticated, able to deal with arbitrary messages (rather than just all-capital-letters ones) and highly resistant to computer attacks. Or at least, to all known attacks after serious attempts by experts to break them (an encryption technique won't be generally adopted until after an extended period of expert attacks) on existing types of computer hardware (new types such as quantum computers will be a different matter if they can be developed sufficiently - some modern encryption techniques are quite vulnerable to them in principle and may well have to be made obsolete in practice in due course).
Gengulphus