Fastest commercially available hardware RNG generator
Linux - HardwareThis forum is for Hardware issues.
Having trouble installing a piece of hardware? Want to know if that peripheral is compatible with Linux?
Welcome to LinuxQuestions.org, a friendly and active Linux Community.
You are currently viewing LQ as a guest. By joining our community you will have the ability to post topics, receive our newsletter, use the advanced search, subscribe to threads and access many other special features. Registration is quick, simple and absolutely free. Join our community today!
Note that registered members see fewer ads, and ContentLink is completely disabled once you log in.
I agree that IBM would be a good choice but Intel provides some good choices. You can get knock offs but then you get what you pay for. It would help to know how you plan to use RNG. Resolution? A simple signal source & A/D will provide a RNG. Remember it relates to the natural signal source as to randomness.
What portion or circuit of the radio are you using to sample in order to produce random number? Sometimes you can use a RC circuit as a filter to pick from the radio circuit and sample throughout the charge cycle thus random. Even better would be zener diode to pick from the RC coupled signal from the sub circuit of your radio generator.
Laser diode circuit would be another choice to produce noise measured with A/D. Really depends on the bit stream necessary and resolution of the sampling circuit.
Well, I'm doing something much simpler, but that still works well. I'm just recording from the radio on an empty channel and compressing it using a hash. Very simple to implement, and it is random according to my tests. It's a good idea to change channels once in a while, and the hash must compress at the right rate or you will get non-random data. The rate was determined experimentally using conventional compression like LZMA2, and random analysis programs like 'ent', dieharder, and I made my own. The sample rate is 48000, and I tested other rates, but they don't work as well because you will either need higher compression or you'll be wasting random data using too high a compression. Basically compressing 48000 at 1/2 will get you 24000 bytes per second of random data. I also wrote a program to feed it to /dev/random, but you can also use rng-tools (which I based it off of, but excluded the ridiculous test).
I would avoid transistors, thermal noise, and diodes, mostly because they may seem random, but may actually be chaotic with long period of repeat ... which I guess is good enough for most applications, but may present a risk. A radio presents a disadvantage as well, if they can guess the channel you are using, or interfere with it. Both are not that plausible because they don't know where you are, and they would need a radio near you to have good accuracy in the prediction. Change the channel often and you're safe from these too.
Last edited by H_TeXMeX_H; 10-15-2012 at 12:18 PM.