The existence of good extensible polynomial lattice rules (Q1408505): Difference between revisions

Basing on the theory of polynomial lattice rules introduced by \textit{P. L'Ecuyer} and \textit{C. Lemieux} in [Modeling uncertainty. An examination of stochastic theory, methods and applications, Kluwer, Boston, 419--474 (2002; Zbl 0979.00016)] and the idea of extensible lattices proposed by \textit{F. J. Hickernell}, \textit{H. S. Hong}, \textit{P. L'Écuyer} and \textit{C. Lemieux} [SIAM J. Sci. Comput. 22, 1117--1138 (2000; Zbl 0974.65004)], the author shows in this paper that for suitable infinite families of polynomial moduli there are generating parameters for extensible rank-1 polynomial lattice rules such that for all these infinitely many moduli and all dimensions \(s\) the star discrepancy and the quantity \(R^{(s)}\) (a measure of digital nets) are small. He considers not only sequences of moduli which are powers of a fixed polynomial of positive degree, but more generally divisibility chains of polynomials. Furthermore, the case of Korobov-type polynomial lattice rules is also considered. All these constructions are useful for quasi-Monte Carlo integration.