New Quantum Entropy Source

Oct 5,  2022

Today we will once again open the curtain on the technological side of Randonautica and tell you about an important event, Randonautica now has its own Quantum Random Number Generator!

As you remember, for the last 3 years Randonautica was receiving quantum entropy from the open API of the Australian National University, but progress does not stand still, so is our research, which imposes more and more requirements on hardware. It was taken up by MMI researcher Scott Wilber, who developed a unique QRNG called PQ128MS, designed to increase the sensitivity of quantum entropy to users' intentions and increase the speed of its generation. But first things first!

So, if you've read the last article about Mind-Matter interaction, then you already know that the generation of Attractor points is based on the ability of the user's intention to influence random quantum processes. The source of such processes is the Quantum Random Number Generator (QRNG). However, among MMI researchers, it is more often called the Random Event Generator (REG), since it gives out not random numbers, but a random series of zeros and ones, as if it were tossing a coin a million times a second and telling you which side it fell.

Many years ago, scientists at Princeton's PEAR Lab noticed that a person's intention, focused on wanting a certain result, can influence the operation of such generators, making the output of ones about 5% more frequent than zeros and vice versa. Later, the first Randonauts figured out how to use this phenomenon to convert such influences into geographic coordinates.

However, not every random number generator is suitable for such experiments, and in this article we will try to explain why.

Random Number Generator Physics

According to their structure, random number generators can be divided into three types: Pseudorandom (PRNG), Physical (TRNG) and Quantum (QRNG). Pseudo-random RNGs create randomness mathematically. There are such RNGs in every computer and they are an algorithm, the input of which is the so-called "seed", that is, the starting point, which can be, for example, the exact time of its launch. From this starting point, the algorithm computes a series of statistically random numbers. However, since this series of numbers is algorithmically created, all numbers in it are predetermined. That is, they are random in terms of statistics, since the probability of each of them is the same, but not random in terms of predestination. For example, if you run such a generator again with the same seed, it will produce exactly the same series of numbers as the first time. Moreover, each such RNG has a certain number of bits, after which the sequence will begin to repeat. Obviously, such RNGs are unsuitable for our experiments, since if all numbers are predetermined by the algorithm, the user's intention cannot have any effect on them.

The second type, physical RNGs (TRNGs), are typically devices that use for randomness generation some physical process that is too complex to be predictable. Such processes are still described by the laws of classical physics, but are no longer considered deterministic due to Chaos Theory.

These include, for example, Thermal Noise, also called Johnson-Nyquist noise. It is electronic noise created by the thermal excitation of charge carriers - usually electrons - in an electrical conductor.

It is not known whether intention can influence such processes, but MMI researcher Scott Wilber believes that such an influence is possible, but will require much more energy than influencing quantum processes.

RNGs are considered quantum (QRNGs) when the Uncertainty Principle applies to physical processes in them and it becomes fundamentally impossible to predict the change in their parameters. For example, when the wave properties of charge carriers begin to change the result of their measurement. This happens, for example, during radioactive decay, the passage of photons through a semi-transparent mirror, or the tunneling of electrons in MOS-transistors, causing shot noise.

Since it is not required to apply a sufficient amount of additional energy to change the value of a bit in purely quantum processes, it is believed that it is easiest to exert a Mind-Matter influence on them. However, in most QRNGs, both quantum and classical physical processes affect the measurement result, and it is almost impossible to separate the entropy produced by them. So, for example, in transistors, along with Shot noise, thermal noise is always present, and the second, as a rule, is much bigger.

Randonautica uses the Quantum RNG from the Australian National University, in which entropy is generated by measuring quantum fluctuations in a vacuum caused by the appearance and disappearance of virtual particles.

However, in the original experiments of the PEAR Princeton Laboratory, which discovered the phenomenon of Mind-Matter Interaction, for these purposes, a different QRNG was used, based on Shot Noise arising from tunneling effects in MOS-transistors. The reason for this noise is also that the current consists of a very large number of individual charges, which makes its flow granular, and the noise itself quantum.

Post-Processing the Noise

A serious problem of physical RNGs is that, unlike pseudo-random noise, physical noise has significant statistical defects. Simply put, this means that sometimes it behaves randomly, and sometimes not. Therefore, in order to make it statistically random, the physical RNG uses a "whitening" procedure, which is also called unbiasing algorithms.

Whitening is usually done by passing statistically imperfect true random numbers through a cryptographic hash function such as SHA-1 or by combining them with pseudo-random numbers through a logical XOR operation.

In the Psyleron QRNG, which was used in the experiments of the PEAR laboratory to eliminate defects, the shot noise was subjected to a logical XOR operation, combined with a pseudo-random noise stream, which reduced the influence of quantum processes on the output data by at least a factor of two.

Most often, in order to increase the amount of output entropy, physical noise is simply fed as a seed to a pseudo-random RNG, thus making its result unpredictable. At the same time, the amount of quantum entropy does not increase, and its influence on each specific bit becomes extremely small.

Our new QRNG

And finally, we move on to the unique QRNG, which has recently appeared on the Randonautica server!

PQ128MS was designed in such a way as to achieve the maximum influence of quantum processes on each bit of the output data and completely get rid of pseudo-random algorithms during correction of statistical defects.

The device is based on ring oscillators. These are closed circuits made of an odd number of transistor inverters that change the signal entering them to the opposite. The signal in such an oscillator oscillates between two voltage levels. Since tunnel leakage occurs in the transistors included in the inverters, Shot noise occurs, which causes jitter in the time of voltage drops and rises. Thus, by reading the signal at regular intervals, certain bit states can be obtained from this jitter.

In the PQ128MS, the circular oscillator has three outputs with an equal number of elements between them, which allows us to get three different signals from it, which are then combined with a logical XOR operation. This triples the probability of sampling a ring oscillator output signal exactly during a transition when the shot noise-induced jitter makes the measurement quantum mechanically indeterminate. Thus, despite the significant influence of thermal noise, quantum Shot Noise creates a contribution that is quite capable of changing the value of the output bit. Further, the entropy from hundreds of such oscillators is combined into a single output using the logical XOR operation, which makes it possible to eliminate statistical defects without using pseudo-random entropy. Thus, all the bits involved in the XOR operation are the product of quantum entropy, which means that its contribution to the result becomes maximum. And finally, outputs from three such generators are XORed together for better reliability.

(read more technically accurate info about the PQ devices in this paper)

After the transition of the Randonautica server to PQ128MS, we expect a significant increase in the impact of users' intent on the process of generating Attractor/Void points. This means that the probability of finding what you think about at these points should increase. Of course, we cannot guarantee this, but that's why it's an experiment.

Also, installing our own QRNG on the Randonavtica server will increase the entropy generation rate to 128Mb/s, which opens up the possibility for us to use new algorithms that require a lot of entropy. We are constantly doing research in our laboratories and perhaps in the near future we will revise the methods of generating coordinates and add bias-amplification algorithms that further enhance the influence of intention on the generation of coordinates. We have already described those methods in more detail in the article "Mind-Matter Interaction", if you have not read it, we recommend reading it.