Re: Matrix-Multiplication: Security?

From: Paul Rubin (//phr.cx_at_NOSPAM.invalid)
Date: 05/27/05


Date: 27 May 2005 05:06:17 -0700

Anton <antonbischof@web.de> writes:
> Ok, may you give me some further information about how "secure" my
> programm is compared with non-linear systems?
> How can I calculate the security of an algorithm?

Nobody knows how to calculate the security of an algorithm. Nobody
knows if there's even such a thing as a secure algorithm. But there
are many algorithms that we know to definitely be not secure.

You might read a book like "The Code Book" by Simon Singh to get an
easy introduction to the subject. For something more mathematically
serious, try:

http://www-cse.ucsd.edu/users/mihir/cse207/classnotes.html



Relevant Pages

  • Re: Meganet on Cryptogram again
    ... > It is only necessary to break the algorithm. ... >> bother to cryptanalyse it. ... If nobody can be bothered to ... > break it, then it is effectively secure, no matter what it's theoretical ...
    (sci.crypt)
  • Re: Meganet on Cryptogram again
    ... > It is only necessary to break the algorithm. ... >> bother to cryptanalyse it. ... If nobody can be bothered to ... > break it, then it is effectively secure, no matter what it's theoretical ...
    (sci.crypt)
  • Re: md5 collisions and speeding tickets
    ... An algorithm is broken if it ... > not have the properties required of a secure hash function. ... at least nobody has provided an example of a successful exploit. ... with real-world information processing. ...
    (sci.crypt)
  • Re: Meganet on Cryptogram again
    ... It is only necessary to break the algorithm. ... > bother to cryptanalyse it. ... If nobody can be bothered to ... break it, then it is effectively secure, no matter what it's theoretical ...
    (sci.crypt)
  • Re: Evaluation of MegaSnakeOil by "expert"
    ... >> unknown to the public. ... Consequently nobody in the ... > it's not such a bad algorithm after all. ... to treat it as a trade secret? ...
    (sci.crypt)