1997-01-27 - Re: OTP security

Header Data

From: paul@fatmans.demon.co.uk
To: Rick Osborne <osborne@gateway.grumman.com>
Message Hash: a5bc4248f63353818485e7c242890c673b6d96d34d21b6844019882235bb89d6
Message ID: <854404295.619789.0@fatmans.demon.co.uk>
Reply To: N/A
UTC Datetime: 1997-01-27 23:44:32 UTC
Raw Date: Mon, 27 Jan 1997 15:44:32 -0800 (PST)

Raw message

From: paul@fatmans.demon.co.uk
Date: Mon, 27 Jan 1997 15:44:32 -0800 (PST)
To: Rick Osborne <osborne@gateway.grumman.com>
Subject: Re: OTP security
Message-ID: <854404295.619789.0@fatmans.demon.co.uk>
MIME-Version: 1.0
Content-Type: text/plain



> I was thinking about the thread we had a week or so ago about OTPs.  Say
> I'm going to burn a CD of what I think are cryptographically random bits,
> but somehow I end up with part of my stream being predictable (say every
> 16th bit).  What does this do to the security of my CD?

Depends on how that 16th bit is related to the other bits and whether 
these predictable bits give any information about the other bits on 
the disk.

If we assume all the other bits are true random and that the 16th 
bits are predictable only in that they can be predicted left and 
right but do not depend upon the other bits not in positions 16, 32, 
48, 16n etc... we can just discard them and use the rest obtaining 
perfect security. We can even use all the bits and all we lose is one 
bit every two bytes and therefore if we are calling the bytes ASCII 
and say adding mod 13 we only have "imperfect" security on every 2nd 
character where there are 2^7 eg. 128 possible characters. 
Suprisingly this yields perfect security as there are still a number of
possible pads which lead to reasonable and plausible decryptions.



  Datacomms Technologies web authoring and data security
       Paul Bradley, Paul@fatmans.demon.co.uk
  Paul@crypto.uk.eu.org, Paul@cryptography.uk.eu.org    
       Http://www.cryptography.home.ml.org/
      Email for PGP public key, ID: 5BBFAEB1
     "Don`t forget to mount a scratch monkey"





Thread