Theory and design of a digital stochastic computer random number generator

The theory and design of a random number generator suitable for use in a general purpose digital stochastic computer is considered. After a review of possible types of random number generator the paper considers the design of a system for the generation of negatively correlated pseudo-random binary sequences. Specifically it is shown that the use of correlated sequences reduces the adverse effects of pure randomness whilst retaining the property of zero correlation between sequences fundamental to the theory of stochastic computation.