Abstract
The paper is devoted to the complexity analysis of binary floating point pseudorandom generators. We start with a stochastic model of a “usual” pseudorandom generator (PRNG). Then integer outputs of this generator are transformed into i.i.d. random variables, agreed with an abstract binary floating point system. Additionally, these random variables are approximately uniformly distributed on the interval [0,1]. Therefore, they can interpreted as (random) outputs of a binary floating point pseudorandom generator (flPRNG). The simulation complexity of such a transformation is defined as the average number of PRNG's outputs necessary to produce the unique output of flPRNG. Several transformations with minimal or approximately minimal complexities are presented and discussed.
References
1 A. Fog, Pseudo random number generators, http://www.agner.org/random/. Search in Google Scholar
2 T. S. Han and M. Hoshi, Interval algorithm for random number generation, IEEE Trans. Inform. Theory 43 (1997), 2, 59–611. 10.1109/18.556116Search in Google Scholar
3 D. Knuth and A. Yao, The complexity of nonuniform random number generation, Algorithms and Complexity, Academic Press, New York (1976), 357–428. Search in Google Scholar
4 D. Romik, Sharp entropy bounds for discrete statistical simulation, Statist. Probab. Lett. 42 (1999), 219–227. 10.1016/S0167-7152(98)00174-6Search in Google Scholar
5 M. Saito and M. Matsumoto, A PRNG specialized in double precision floating point numbers using an affine transition, Monte Carlo and Quasi-Monte Carlo Methods 2008, Springer, Berlin (2009), 589–602. 10.1007/978-3-642-04107-5_38Search in Google Scholar
6 N. Vorobjeva, A. Korobeinikov and V. Nekrutkin, Optimal simulation of discrete distributions (in Russian), Vestnik St. Petersburg Univ. Math. 1 (2012), 3, 14–23. Search in Google Scholar
7 IEEE Standard for floating-foint arithmetic, IEEE Std 754-2008. Search in Google Scholar
© 2016 by De Gruyter
