From: Timothy Newsham <newsham@wiliki.eng.hawaii.edu>
Date: Mon, 21 Jun 93 20:40:44 PDT
To: zane@genesis.mcs.com (Sameer)
Subject: Re: OTP dual decryption
In-Reply-To: <m0o7qY1-000MU6C@genesis.mcs.com>
Message-ID: <9306220340.AA02067@toad.com>
MIME-Version: 1.0
Content-Type: text/plain


> 
> 	Hmm? I don't understand this problem. There's only one set of
> cyphertext.. the actual cyphertext. Do you mean "different batch of
> cyphertext" as the OTP which creates the innocuous plaintext from the
> cyphertext? Maybe encrypt the OTP w/DES and keep it on your hard drive.
> When "they" snag the drive, they see the different cyphertext, you tell
> them that it's the OTP you used and give them the DES-key to decrypt the
> innocuous OTP.
> 	I sense a problem with histogram equalization, however. Is there a
> problem here or does OTP-encryption take care of that? 

With a OTP any plaintext can correspond with any cyphertext
given:

   PAD (+) plaintext = cyphertext
meaning
   PAD (+) cyphertext = plaintext
and
   plaintext (+) cyphertext = PAD

so, take your output of DES,  xor its contents with
desired "false plaintext"  this is your false pad.
store this on a seperate disk and make it look
all secret-like.

When feds come and take your cyphertext (which is output
of DES) they ask for your key.  You hand over your
disk with (fake) PAD.  They xor the two together to
get (fake) plaintext which reads.  

You dont have to do this before hand either, if you
have a copy of your cyphertext after the feds
confiscate a copy from you you can generate pad
to make it say something that will really freak them
out,  like something you couldnt have known prior
to the seizure of the cyphertext :
  "I will be illegally raided on July 2, 1994,  sounds
   sort of like '1984' to me.  I wish the government
   would stay out of my  life"


> | Sameer Parekh-zane@genesis.MCS.COM-PFA related mail to pfa@genesis.MCS.COM |

peace.