Randomised Decision Making and Cyber Security

Randomizing the order of packet transmission or paths taken by data can help in preventing interception and securing communications.
The key advantage of randomness is its unpredictability, which makes it difficult for others to exploit patterns or anticipate actions.

The artistic illustration above demonstrates the concept of multiple data paths in a computer network. It shows how data can travel through various routes between nodes like servers, routers, and computers. This complexity and the use of different paths enhance security by making the data transmission less predictable and more resilient against interceptions or attacks.

Physical processes used for generating true random numbers typically involve measuring some aspect of a naturally occurring, inherently unpredictable phenomenon. These include:

1.  Radioactive Decay: The decay of radioactive elements is fundamentally random at the atomic level. By measuring the time between decay events, truly random numbers can be generated.
2.  Thermal Noise: Also known as Johnson-Nyquist noise, this is the electrical noise generated by the thermal agitation of charge carriers (usually electrons) inside an electrical conductor at equilibrium. This can be measured and converted into a random number.
3.  Photonic Emission: Measuring the quantum fluctuations in the emission of photons from a light source can provide random data.
4.  Atmospheric Noise: Radio or sound signals from the environment, which are essentially random due to various unpredictable atmospheric phenomena, can be used.
5.  Quantum Phenomena: Various quantum processes, like quantum tunneling in semiconductors, can be observed and translated into random numbers.

These methods are used in high-stakes security applications like cryptographic key generation, where unpredictability is crucial. The hardware devices that use such phenomena to generate random numbers are known as True Random Number Generators (TRNGs). Unlike PRNGs, TRNGs do not rely on a predetermined algorithm, making them more suitable for tasks where high security and unpredictability are required.