Quantum upgrade makes random number generation fully traceable

[Alfonso Maruccia]

Trust, but verify: Random number generation is a serious matter in modern computing. Most systems rely on a purely hardware-based approach to RNG, but the process is essentially impossible to verify or trace. Now, a new quantum-based method developed by US researchers could offer a potential solution to this RNG puzzle.

Physicists at the National Institute of Standards and Technology in Boulder, Colorado, have upgraded their previously developed quantum-based method for true random number generation. Thanks to the addition of a blockchain, the process can now be considered fully transparent and traceable.

The NIST team first introduced their RNG technique in 2018, exploiting quantum entanglement between pairs of subatomic particles. What Albert Einstein famously called "spooky action at a distance" describes how two entangled particles share the same quantum state regardless of the distance between them, meaning neither particle can be described independently of the other.

The NIST device generates pairs of entangled photons, the elementary particles that carry light, radio waves, and other forms of electromagnetic radiation in our universe. The photons are then sent to two separate measuring stations 90 meters apart, where their polarizations are detected and converted into a series of digital bits.

Must read: Number Representations in Computer Hardware

In a newly published study, NIST scientists describe a random number generator that is fully traceable. Their previous method was enhanced by adding a publicly verifiable blockchain log, where hashes are recorded through a NIST-developed protocol called Twine. The team has turned this new research into a public service named CURBy (University of Colorado Randomness Beacon), designed to provide a periodic source of randomly generated strings.

The study explains that while algorithms used to generate pseudo-random numbers are typically auditable, they cannot guarantee that their output is truly unpredictable – even to someone who knows the original seed. The new protocol extracts randomness from the quantum correlations in the original device and uses hash chains to cryptographically verify the extraction process.

Third parties can access the blockchain's public data to verify whether the RNG process has been tampered with. The NIST team also achieved significant improvements in RNG speed and efficiency: the latest experiment produced 512 bits of random numbers in about 20 seconds, compared to 10 minutes for the same output in 2018.

According to Peter Brown, associate professor at France's Institut Polytechnique de Paris, a fully traceable system should provide an additional layer of trustworthiness to the entire RNG process. "In the current era of misinformation, at least we will be able to trust in randomness," Brown said.

Permalink to story:

Quantum upgrade makes random number generation fully traceable

 

[maxxcool7421]

Since ultimately nothing is truly random that tracks ..

 

[Xelions]

I can predict each lotto number at just before each number is drawn but it doesn't benefit me if I can't get a coupon on it

I'll stick to the Akashic record thanks

 

[unoficialoficial]

That method which using quantum entanglement does not produce true random data. To generate true random data (like numbers), you can't simply create them; you need to extract them from a source with high entropy. The high entropy of the source guarantees the true randomness of the extracted data. A high-entropy source should have a chaotic function or mechanism. Quantum entanglement isn't chaotic, so it doesn't produce true randomness because it lacks high entropy. It's uncontrollable, but that doesn't mean it has high entropy. The inability to interfere with the mechanism doesn't automatically guarantee randomness. The mechanism's *inherent nature* is what matters. Interference ability is irrelevant, we just need an output clean from statistical patterns.
True random data don't have patterns inside them, that's the only criterion. People often make mistake because they don’t spend too much time philosophically approaching the concept of true randomness. They confuse the uncontrollable(without ability for interference) with the unpredictable(without statistical patterns).

So now that we have a solid understanding of what true random data means and what we need, let's build an algorithm to generate true random data. We need a high-entropy source with an underlying chaotic function. For this, we can use the SHA algorithm! The SHA algorithm produces a hash output string with vast changes even with tiny changes in the input string, making it a chaotic function and a source of high entropy. So that would be our source from which we will extract the true randomness.
For our algorithm, we'll take the current time string with millisecond accuracy add the RAM usage string in byte accuracy and optionally the pointer pixel as a string. We'll then calculate a SHA1 hash (to apply first layer of chaos and inflate the seed) on this combined string, followed by a SHA256 hash on that SHA1 result. This SHA256 hash will be our true random string. So if for example we need a true random number between 0 and 999, we can take the modulo of that hash256 converting it into a single large integer and then calculate largeinteger % 1000.

This practical method/algorithm generates true random numbers(I think it’s the first algorithm which produce true random data so I give it ZLIB license as the inventor, so you cannot patent it ) and takes less than half a second on a digital computer. It's available to everyone with a CPU. The method using the quantum entanglement phenomenon does not produce true random numbers and of course is not practical and easy accesible.