On the Use of Reducible Polynomials as Random Number Generators
- 1 January 1993
- journal article
- Published by JSTOR in Mathematics of Computation
- Vol. 60 (201) , 363-374
- https://doi.org/10.2307/2153173
Abstract
The randomness properties and the hierarchy of correlation coefficients are studied of approximate-maximum-length sequences, for which the characteristic polynomial is a product of several primitive polynomials. The randomness properties are almost the same as for maximum-length sequences characterized by a primitive polynomial with many terms and of the same degree. Reducible characteristic polynomials have acceptable figures of merit and can be of extremely high degree. Since they are also easily constructed and implemented, reducible polynomials are strong candidates for reliable random number generation, especially at the bit rates needed in large-scale Monte Carlo simulations.Keywords
This publication has 9 references indexed in Scilit:
- Shift Register SequencesPublished by World Scientific Pub Co Pte Ltd ,2014
- Recent trends in random number and random vector generationAnnals of Operations Research, 1991
- The hierarchy of correlations in random binary sequencesJournal of Statistical Physics, 1991
- A review of pseudorandom number generatorsComputer Physics Communications, 1990
- Figures of Merit for Digital Multistep Pseudorandom NumbersMathematics of Computation, 1990
- Maximum-length sequences, cellular automata, and random numbersJournal of Computational Physics, 1987
- Optimal characteristic polynomials for digital multistep pseudorandom numbersComputing, 1987
- Property of cosets and its application to the calculation of moments of binary sequences having multipliersElectronics Letters, 1971
- Primitive trinomials whose degree is a Mersenne exponentInformation and Control, 1969