Path: cactus.org!milano!cs.utexas.edu!zaphod.mps.ohio-state.edu!mips!swrinde!
+     gatech!bloom-beacon!eru.mt.luth.se!lunic!sunic2!mcsun!sun4nl!hacktic!
+     bill
From: bill@hacktic.nl (Bill Squire)
Newsgroups: sci.crypt

Subject: Re: generating one-time pads
Keywords: one-time pad, generation of
Message-ID: 
Date: 18 Jun 92 16:45:17 GMT
References: 
Organization: Hack-Tic Magazine
Lines: 26

em@topgun.ucsb.edu (Mehlschau;Ed;;;;CCSE;CCSE-Staff;990630;;;CCSE-Operations;500  writes:

> Are there any products available that generate one-time pads, in
> particular, arbitrarily long sequences of random bits that are the
> result of some apparently random process like radioactive decay?  At
> what rate can they pump out the bits?  How do they work?  How much do
> they cost?
> 
> No jokes about audio DAC's connected to radios tuned to unused
> parts of the spectrum please...  :->
> 
Whats wrong with that method?  Sounds alot more random than any computer
program doing it.  If you disconnect the antenna and short it, it'll be
even more random.  A similar method involves using a lightpen on a TV tuned
to a vacant channel.

A method I found to work quite well was to produce diode white noise,
highpass filter it at a rather high frequency (above 100kHz to avoid
50/60Hz disturbance) pass it thru a toggle flipflop and use!  The
other side of the D-flop can be used to clock the truely "random" data
stream.  Placing the whole device in a well shielded RF box and bringing
out the bits with fibre, will result in the second best thing to using the
decay of "something radioactive", without the objections of having "something
radioactive" at your home or work.

                                                        Bill