home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: sci.crypt
- Path: sparky!uunet!spool.mu.edu!agate!linus!linus.mitre.org!gauss!bs
- From: bs@gauss.mitre.org (Robert D. Silverman)
- Subject: Re: Fraction part of squareroots as one time pads ?
- Message-ID: <1993Jan11.143125.24040@linus.mitre.org>
- Sender: news@linus.mitre.org (News Service)
- Nntp-Posting-Host: gauss.mitre.org
- Organization: Research Computer Facility, MITRE Corporation, Bedford, MA
- References: <1993Jan11.111351.7777@hsr.no>
- Date: Mon, 11 Jan 1993 14:31:25 GMT
- Lines: 41
-
- In article <1993Jan11.111351.7777@hsr.no> frank@hsr.no (Frank A Stevenson) writes:
- :At one time I calculated pi and e with several houndred thousand decimal digits,
- :and after taking interest in cryptology the question sprung to mind: Would it be
- :safe to use the binary expansion of an irrational number as a one time pad. To my
- :knowledge suche sequences of bits exhibits no knowns statistical patterns, and
- :provided the key is large enough >60 bits, the one time pad would be sufficently
- :hard to find.
-
- Let me first politely ask that you limit the length of your lines. Run-over
- lines are hard to read.
-
-
- Your last point might be questioned. There are some good integer relation
- finding algorithms (Ferguson & Forcade, for example and improvements)
- that allow one to very quickly determine that the number is the root of
- an integer, based upon just the first few digits. Once that is determined,
- finding the rest of the key is trivial.
-
- The same would be true for any ALGEBRAIC irrational.
-
- :
- :If I where to use a 68000 processor for the implementation of this scheme two
- :difficulties arises:
- :
- :1) Taking square roots is rather expensive, and this limits the size of the
- :message to be encrypted.
-
- Define "expensive". Using FFT multiplies, once could compute a square
- root of an integer to several thousand bits in just a few seconds on a
- modern workstation.
-
- :
- :2) For effeciency reasons it is best to extract square roots limited to 32 bits
- :of length. (The greatest divisor fopr the DIVU command) This provides a rather
-
- Why is this? Use multi-precision techniques.
- --
- Bob Silverman
- These are my opinions and not MITRE's.
- Mitre Corporation, Bedford, MA 01730
- "You can lead a horse's ass to knowledge, but you can't make him think"
-