What to do when pseudo- is not good enough
?
1 like?533 views
This document discusses various options for generating true random numbers when pseudo-random numbers are not sufficient. The options include: 1. Buying a true random number generator USB device, which is expensive for empirical tests. 2. Using the service Random.org, which generates numbers from radio static, but is free only for sequences up to 10,000 numbers. 3. Manually inputting numbers from "A Million Random Digits", which would be time-consuming. It also briefly outlines less viable options like fabricating a random number generator, using a webcam or thermal noise, or a lava lamp and webcam software. Overall, the document explores both inexpensive and more sophisticated options for obtaining true random
What to do when pseudo- is not good enough
- 3. What to do when PSEUDO- is not good enough? (this message has little to do with scientific results of its author but is actually inspired by the many questions he has received about a topic that he just happens to know by chance*) rand() true randomness * at the same time the author is very comfortable with the pseudo-randomness for most of the way...
- 4. 1. Buy a True random numbers generator (USB device) http://www.westphal-electronic.com/ upside: It is a very sexy idea of having true randomness in your pocket.. downside: it is kind of expensive for several empirical tests
- 5. 2. Go to Random.org http://www.random.org/ The owner there generates true random number by perceiving the static of tree radios. You should read carefully most of the website before deciding to use the numbers. Pay special attention to ＾Introduction￣, ＾Statistics￣ and ＾FAQ￣. upside: - everyone can just download real random numbers over the WWW - you can use several ready-made scripts http://basaga.hit.bg/RANDOM_MATLAB.zip http://basaga.hit.bg/truerand.zip downside: - it is free for sequences up to 10000 numbers
- 6. 3. You can always use the classical approach: upside: maybe it is the final prove that you are nostalgic downside: Somebody will have to input ＾A million random digits￣ by hand
- 7. 4. If you are good at fabricating http://robseward.com/itp/adv_tech/random_generator/ upside: there you go C now you are creating randomness downside: you will probably have to improve your precise bonding skills
- 8. 5. Buy a webcam. ＾Buy one of those $25 toy digital cameras. Keep the lens cap on, or put black tape over the lens. Connect to USB port. Add script to snap a "picture" every few minutes to prng. With no light, digital cameras return thermal noise - which looks like "snow" on an analog TV. I've done this with a toy camera I bought for my daughter. The camera feeds raw pixels to the linux driver, and the post processing done by the Windows software was never implemented in Linux, making it useless as a camera (plus it has 256M ram, but no flash memory). But it works great for this application. I haven't done a mathematical analysis of exactly how much entropy is in the signal. I'll leave that for the stat geeks.￣ (http://it.slashdot.org/article.pl?sid=07/11/12/1528211) Note: It should work with any digital camera as long as you have the means (cable) to input and snap (script) its signal into an application. It is recommended to set the ISO sensitivity as high as possible as a way to amplify the entropy upside: CHEAP downside: You should do a lot of testing before using that.
- 9. 6. Lava lamp. You actually buy (or own anyway) a lava lamp AND a webcam. Then you download the software LavaRnd (http://www.lavarnd.org) and follow the instructions there upside: You have a beautiful decoration at your working place. downside: after so many tries you don¨t really care about the true randomness anymore