Researchers have proposed a new random number generator (RNG) design that signals a shift away from traditional methods. So, let’s take a look at what RNGs are, what this innovation means, and how this impacts its associated technologies.
What Are RNGs?
At their most basic, RNGs generate random, non-sequential numbers, letters, characters, or associated programmable symbols. In other words, they digitally mimic the essence of luck. For this reason, they are most commonly used in the iGaming industry, as they can ensure that the results of digital slot games are as random as pulling a lever and spinning the reels. For example, players who try the Starburst slot will find that a variety of different coloured gems will fall onto the reels in a random order, as well as the star wild which activates the bonus feature of the game. By using RNGs, this ensures that the results of each spin are entirely down to chance, making the gameplay as fair and immersive as possible.
Elsewhere, the technology can also be leveraged for online security protocols. RNGs are one of the simplest ways to create secure passwords, as they can generate a string of unrelated letters, numbers, and special characters that are not rooted in reality. As they are not predetermined, based on your personal data, or sequential, this makes them harder for malicious attacks to crack. Taking this to another level, RNGs are the key technology behind cryptography, as they generate unpredictable values that maintain the security of cryptographic systems, such as initial vectors, encryption and decryption keys, and nonces.
Breakthrough in RNG Methodology
Typically, there are two types of RNG – those which use software, and those which use hardware. Of the hardware-based RNGs, perhaps the most common type utilises the multiplicity of quantum mechanics (QRNGs). QRNGs have seen a wealth of innovation and development over recent years, from the experimentation of materials like perovskite in the entropy source LEDs, to the shrinking size of such devices into chips. Most recently, researchers at Toshiba Europe developed a QRNG chip that can be integrated directly into devices which is capable of generating secure and reliable random numbers at a rate of 2 Gbit s-1.
Taking a more unusual approach, however, a team from the University of Science and Technology of China (USTC) of the Chinese Academy of Sciences (CAS) proposed a design methodology for RNGs that use additive white Gaussian noise (AWGN) as their source of entropy. The study, which was published by IEEE, outlines the design for flexible Gaussian RNGs (GRNGs) with a reconfigurable σ variance and scaling index, which is capable of a theoretical output range of ±14σ.
Impact of GRNG Innovation
The innovative GRNG design is specifically tailored to SerDes simulation systems – a functional block used in chip-to-chip or point-to-point communication that has the ability to serialise and deserialise digital data. One of the downfalls of the current solutions is that they generally need lots of multipliers and rounding units, which not only uses a lot of power and energy, but can lead to higher error rates.
This proposed GRNG, however, theoretically would improve the performance of SerDes systems, boosting resource utilisation, raising clock speeds, and improving degrees of parallelism. The GRNG itself also boasts better flexibility, easier reconfiguration, increased ranges for values and outputs, and more stability. With this in mind, this novel GRNG design could flip the script on the reliability and security of these systems, as well as paving the way for a new generation of RNGs.
