1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Help with a new cipher

  1. Aug 29, 2010 #1
    Last edited: Aug 29, 2010
  2. jcsd
  3. Aug 29, 2010 #2


    User Avatar
    Science Advisor
    Homework Helper

    I don't think the key improves security. It should not be too difficult to look at the fractional parts and decide what numbers are possible keys (depending on how many digits are given). And depending on how many digits you give, you may also clip letters from the end of long words.
  4. Aug 31, 2010 #3
    what if the division gives irrational number?
  5. Aug 31, 2010 #4


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    What if dividing two integers gives an irrational number?

    I dare say you probably don't care what the message is anymore.

    The simplest attack on this is obviously the fact that you need a key.

    A lot of work might be able to be done to determine what the key is even if somebody doesn't know. For example, if I see that one of your encrypted words is .015 I know that your key must be relatively large, so starts with a letter near the end of the alphabet. I'm sure that some number theory could be used to drive home at possible keys.

    There's no real reason to break up your numbers into blocks of 8 besides this cell phone thing. Usually characters are broken up into blocks so that when transmitted you don't give away what the word lengths are. You don't have such a problem here, and it could make it harder to decrypt if you use different length codewords, because then you'll be transmitting a wider range of numbers. This would mean that the intercepting party has no idea how long the codeword is which would probably be a good thing

    One good thing is to not send over the whole decimal, which is how you presented it. Otherwise I could simply find what the decimal is in smallest form and then just scale it upwards until the numerator and denominator both fit the pattern necessary for a codeword. I tried that on two of the decimal numbers you had and it doesn't work because you don't get the full number
    Last edited: Aug 31, 2010
  6. Aug 31, 2010 #5


    User Avatar
    Science Advisor

    My initial thought is that the security of modern ciphers is such that knowledge of the algorithm, which is typically publicly available, does not jeopardize the security of the cipher. In your case, if I know the algorithm, a simple brute force attack on the key space should crack the cipher, since I need only to multiply each cipher text word with different keys. Now, the security of your cipher certainly improves as you make the key longer. However, now you must also make the world length longer. In the limit that the key becomes equal in length to the plain text message, you have the well known one-time pad, which is indeed provably secure (in the event that your key is truly random and used only once).
  7. Aug 31, 2010 #6


    Staff: Mentor

    That can't possibly happen. Both numbers involved in the division are integers, so the quotient will necessarily be rational. By definition, a rational number is the quotient of two integers.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook