An Empirical Study of Non-Cryptographically Secure Pseudorandom Number Generators

This paper gives an empirical study of various pseudorandom number generators (PRNG), along with their brief history, from the traditional ones such as Middle Square Method to the most advanced ones like Xoroshiro128+, along with various modifications that were added to them. Comparing them on the basis of their computational times, code complexities and the algorithms used by them, it can be deduced that which of the PRNG is the fastest, or which gives the most equidistant random numbers and which generator is most useful for a particular purpose. This paper doesn’t cover any cryptographically secure pseudorandom number generator (CSPRNG) and all the generators mentioned henceforth are cryptographically insecure.