Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!rutgers!mit-eddie!mit-amt!simsong
From: simsong@mit-amt.MEDIA.MIT.EDU (Simson L. Garfinkel)
Newsgroups: sci.math,sci.physics,sci.crypt,sci.math.stat
Subject: Re: Do you use RANDOM NUMBERS?
Message-ID: <836@mit-amt.MEDIA.MIT.EDU>
Date: Wed, 18-Mar-87 10:25:38 EST
Article-I.D.: mit-amt.836
Posted: Wed Mar 18 10:25:38 1987
Date-Received: Fri, 20-Mar-87 00:37:16 EST
References: <5712@reed.UUCP>
Reply-To: simsong@media-lab.MEDIA.MIT.EDU (Simson L. Garfinkel)
Organization: MIT Media Lab, Cambridge MA
Lines: 28
Keywords: Random numbers, leprechauns
Xref: mnetor sci.math:745 sci.physics:927 sci.crypt:253 sci.math.stat:69

In article <5712@reed.UUCP> mdr@reed.UUCP (Mike Rutenberg) writes:
>How do people currently use RANDOM NUMBERS?  Simulations?  Anything else?

A few years ago I was doing research in PSI (no joke) and needed a 
truely random source of numbers. Not wanting to get involved with
radioactive decay (and for other reasons wanting to use electronic
circuits) I built an electronic random number generator which was
then attached to a single bit of a digital input port on an IBM-PC.

The random number generator consisted of a variable-duration pulse generator.
When the pulse was long (past a certain point), a comparator output a "1"
which was fed to the I/O port and fed back to the pulse generator to shorten
the duration of the pulse. Likewise, when the output was a "0" the duration
of the pulse was lengthened.

Extensive FFT analysis of the output of the generator showed it to have
roughly uniform frequency distriubtuion with the exception of a high
energy point around 100KHZ. As the sampling rate was under 10HZ, this did
not present a major problem.

Incidently, I didn't discover any PSI effects.

				Simson L. Garfinkel

-- 
---
Overheard in a terminal room: "What an interesting pattern that is on
your shirt. May I trace it with my finger?"i