Ultrafast physical generation of random numbers using hybrid Boolean networks

Area of Science:

• Physics
• Computer Engineering
• Information Theory

Background:

• Physical random number generators (PRNGs) are crucial for secure communication and simulations.
• Existing PRNGs often require complex postprocessing or have limited throughput.
• Boolean networks offer a promising framework for chaotic signal generation.

Purpose of the Study:

• To design and implement a high-speed physical random number generator.
• To leverage hybrid Boolean networks for efficient chaos generation.
• To achieve high-throughput, postprocessing-free random number generation.

Main Methods:

• A hybrid Boolean network architecture was designed using autonomous and clocked logic gates.
• The network was implemented on a reconfigurable chip with a bidirectional ring topology for autonomous gates.
• The output of autonomous gates drove clocked gates to produce random numbers.

Main Results:

• The generator demonstrated broadband chaos generation from autonomous logic gates.
• Physically generated random numbers passed standard randomness tests without postprocessing.
• An implementation of 128 generators achieved a real-time bit rate of 12.8 Gbits/s.

Conclusions:

• The hybrid Boolean network approach enables high-speed, high-quality physical random number generation.
• Reconfigurable chip technology facilitates parallelization for massive throughput.
• This method offers a robust solution for applications requiring large volumes of random numbers.