Number Generator Research and Development Project
Cryptography · Security · Numerical Methods · Physical Modeling · Artificial Intelligence
Random Interval Paradigms
The classic text, A Million Random Digits with 100,000 Normal Deviates is available on-line.
Intel has posted
a considerable amount of information related to the Intel RNG in the 8xx
series chipsets (including the 850 chipset for the new Pentium 4), Intel
security drivers which utilize the RNG, and articles concerning the
necessity of truly random number generation for information security.
The following links can be found through the above link and are provided here for convenience:
link cannot easily be found through the above links:
For additional information on the Intel RNG, see U.S. Patent Documents:
I. Random number generation based on quantum-mechanical uncertainty in spontaneous nuclear reactions provided by timing intervals between geiger-counter signals. Similar to RAND project and Intel RNG inasmuch as all generate random numbers by timing random intervals.
II. Bias reduction by modular addition.
Silicon Graphic's Lavarand™
The lighter side of random number generation: Generating random numbers by seeding a pseudo-random generator with a hash of a bitmap produced by a digital camera aimed at a set of Lavalite lamps. See U.S. Patent Document:
Generating random numbers by seeding a pseudo-random generator with a hashed video image. See U.S. Patent Document:
Generating random numbers by seeding a pseudo-random generator with a hash of system parameters. See U.S. Patent Document:
Tests for Randomness
Processing Standards Publication 140-1 (FIPS PUB 140-1) specifies certain
tests for randomness to be performed on truly random number generators
(RNGs) used in information security applications.
Diehard is a battery of tests for random number generators developed by Dr. George Marsaglia of Florida State University Department of Statistics. Originally developed for testing pseudo-random generators, Diehard has since become a de facto standard for testing RNGs. Diehard is much more rigorous than FIPS so that, as recently as 1997, Marsaglia himself stated, "I have found none of the latter [truly random number generators] that get past DIEHARD." We are aware of two: the Intel RNG and our xRNG.
To determine if an RNG "gets past Diehard" it is important to pay particular attention to exactly what are the pass/fail criteria. The following guidance is provided by Marsaglia:
you should not be surprised with occasional p-values near 0 or 1, such as .0012 or .9983. When a bit stream really FAILS BIG, you will get p's of 0 or 1 to six or more places. By all means, do not, as a Statistician might, think that a p < .025 or p> .975 means that the RNG has "failed the test at the .05 level". Such p's happen among the hundreds that DIEHARD produces, even with good RNG's.
Diehard requires approximately 10 megabytes of random numbers.
xRNG technology is protected by US Patent. Other foreign and domestic patent rights may be pending in this technology. No license to or transfer of any intellectual property rights are granted expressly or by implication, estoppel, or otherwise by this document. All brands, names, and trademarks are the property of their respective owners. Links to other sites are provided as a courtesy. RNG Research disclaims any responsibility for the condition of such links which are provided "as-is." We will be adding additional references in the upcoming weeks. We welcome any suggestions.
Copyright © 2000-2006