Analysis of Mixing of Pseudo – Random Number Generators with Statistical Tests

Pseudorandom number generators (PRNGs) are essential in the safe and authentic simulation incepts by generating seemingly unbreakably unpredictable sequences. Nonetheless, these single PRNG systems still have certain design flaws, and even basic human analysis can uncover patterns that are not random which lowers the overall entropy value of anything they base outputs on. In the current paper, we present a new method that solves these problems by increasing the PRNGs unpredictability with no dependence on other randomness sources. The motor behind enhancing the entropy and reliability is based on the construction of compound PRNG systems. Various tests are performed on the model implementations with different mixing procedures, and incorporating the Frequency Test and the Run Test, the Matrix Rank Test, and others, illustrates that this method improves the operation of PRNG based systems increases the overall security and reliability of these systems against predictability and low entropy levels.