I have been reading various sources for experimentation with truly random numbers. As I understand, it is impossible for a computer to generate a "truly random" number as they are deterministic in nature (easily predictable).
One of the methods I have seen in some embedded hardware random-number generators is through measurements of radioactive decay, which is arguably a "truly random" noise source. I have a small DIY Geiger counter that I have salvaged from a separate project that I can use to measure ionizing radiation. This device can communicate via an interrupt pin on devices such as the Arduino or similar.
One likely flawed implementation I have quickly come up with follows this set of steps:
- Increment X from 0 to 2^32 (max unsigned integer) in a fixed Y delay
- Once the Geiger interrupt is sent the Arduino, capture X
- Apply modulo 2 to X (giving a binary 1 or 0)
- Repeat until there are a desired number of bits (let's say 8 bits for example)
My main question is, am I essentially eliminating randomness by relying on software (a deterministic device) for these tasks? By following these specific set of instructions for generation, am I making the numbers deterministic? How are random numbers measured for entropy and are there any standards for doing so?
Worth noting: Nothing I am experimenting with is mission critical, just for fun. This involves several concepts from an assortment of subjects (Math, physics, programming, etc) so I'm really not sure if this belongs here.
Thanks
I am also doing some research in TRNG and entropy. I am using a TRNG Device from protego , which uses a Zenner diode and an amplifier to generate the random bytes.
Regarding entropy, I came to the conclusion that Shannon Entropy is the preferred way to compute entropy. I hope this will help you a little bit.