(1) Sabanci University, Orhanli, 34956 Tuzla, Istanbul, Turkey;(2) Johann Radon Institute for Computational and Applied Mathematics, Austrian Academy of Sciences, Altenbergerstr. 69, 4040 Linz, Austria
Abstract:
Nonlinear congruential methods are attractive alternatives to the classical linear congruential method for pseudorandom number
generation. Generators of higher orders are of interest since they admit longer periods. We obtain lower bounds on the linear
complexity profile of nonlinear pseudorandom number generators of higher orders. The results have applications in cryptography
and in quasi-Monte Carlo methods.