Random Number Generation

Question

I have a need for true random number generation. I've been looking at a number of hardware solutions. Does anybody have any solid experiences with these? Any recommendations?

I don't think I can make a convincing argument for a lava lamp hack. ;)

Not that I don't think there will be some good answers here, but SO's tag on the subject has some goodies http://stackoverflow.com/questions/658622/true-not-pseudo-random-number-generators-whats-out-there — Kara Marfia, Jun 18 '09 at 20:15
If you can't convince them that the lava lamp hack is not awesome, then you're obviously working for the wrong people. ;-) — MathewC, Jun 18 '09 at 20:23
Bah, I should have checked stackoverflow as well. I've toyed with the LavaRND solution, but I don't think the lavalamp in the DC will fly. ;) It'd make for some fun nights, though. — dr.pooter, Jun 18 '09 at 20:27

score 4 · Answer 1 · answered Jun 18 '09 at 20:13

4

Is there such a thing as a true random number generator? Random.org uses atmospheric noise to generate a random number, just need a really really long antenna.

answered Jun 18 '09 at 20:13

Nate

524
3
10

score 2 · Answer 2 · answered Jun 18 '09 at 20:14

2

Here is one idea:

"True random numbers from Wi-Fi background noise"

http://www.freewebs.com/pmutaf/iwrandom.html

answered Jun 18 '09 at 20:14

Joseph

3,787
26
33

Great link. I'd missed this in my googling apparently. – dr.pooter Jun 18 '09 at 20:24

score 1 · Accepted Answer · answered Jun 18 '09 at 20:18

1

We once bought a Protego SG100, but they hadn't released the Linux driver when we needed it so we never got around to using it. It looks like they have some newer products with various drivers now.

answered Jun 18 '09 at 20:18

Gene Gotimer

2,442
20
16

Very cool. I've taken a look at the Protegos and Linux support seems to be there, but I'm not sure on reliability. I've also been looking at a solution from Comscire. Thanks! – dr.pooter Jun 18 '09 at 20:22
This was 10 years ago. It looks like they have Linux support now. – Gene Gotimer Jun 18 '09 at 20:33

score 1 · Answer 4 · answered Jun 18 '09 at 20:38

1

Unless you need hundreds of billions of random #s, the easiest thing to do is just use random.org. You can purchase blocks from them if you go over the free allotment.

People have come up with some rather unique solutions to gain "street cred" with their random # generators (particularly online gaming sites). This random dice roller is my personal favorite ... generates 1.3 Million random dice rolls per day.

answered Jun 18 '09 at 20:38

Beep beep

1,843
2
18
33

+1 for the dice roller, I was going to post that haha – Element Jun 19 '09 at 00:48

score 1 · Answer 5 · answered Jun 18 '09 at 20:41

If you want some scientific background here is the Random Number Generator homepage of the at the Department of Mathematics of the University of Salzburg in Austria.

We present results and links for this fundamental tool in stochastic simulation and in applied cryptography, some of them due to our own research in this field. Enjoy the data and allow for necessary incompleteness and subjectivity.

Here is another interesting page about REG design in a project at Princton University:

The Global Consciousness Project uses three different random event generators (REG or RNG). These are the PEAR portable REG, the Mindsong Microreg, and the Orion RNG. All three use quantum-indeterminate electronic noise.

They are designed for research applications and are widely used in laboratory experiments. They are subjected to calibration procedures based on large samples, typically a million or more trials, each the sum of 200 bits. In the GCP application, an unbiased mean is guaranteed by XOR logic. Although they have different fundamental noise sources, they all provide high-quality random sequences that are functionally equivalent.

score 1 · Answer 6 · answered Jun 18 '09 at 20:52

The Via C3 processor has a hardware random number generator based on two units built from several . Drivers are available in the Linux kernel.

There's a detailed analysis available and the results seem quite favorable. It wouldn't be hard or expensive to buy a few C3 boards and rack them in your DC.

score 1 · Answer 7 · answered Jun 18 '09 at 21:45

Your question makes me think of the truerand library from... what was it, 1999? The idea there was getting true randomness from measuring clock drift. If I recall correctly, the idea was that that's subject to quantum phenomena, so whatever entropy it gives you is really, really good entropy.

score 1 · Answer 8 · answered Jun 18 '09 at 21:48

Take a look at http://www.fourmilab.ch/onetime/. It's a one-time pad generator package that contains the source code to a separate, software-based pRNG. At the very least, it will provide some insight on the complexities of generating randomness on a computer.

score 0 · Answer 9 · answered Jun 18 '09 at 22:13

0

Applied brownian motion or weibull distribution.

answered Jun 18 '09 at 22:13

Niklas Rosencrantz

169
2
17

IANAMM (I am not a math major). But I play one on ServerFault. To the best of my knowledge, each of these has deficiencies when considering true randomness. Right? – dr.pooter Jun 18 '09 at 22:29
7 generators are Lehmer, Rotenberg, GGL, Neave Oakenfull (2) and Wichmann-Hill docs.python.org/library/random.html Weibull function or Brownian motion are also wellknown randomnesses: import random print random.weibullvariate(2,2) output: 1.85255758863 – Niklas Rosencrantz Nov 29 '09 at 05:54

Random Number Generation

9 Answers9