Stationary self-similar random fields on the integer lattice
We establish several methods for constructing stationary self-similar random fields (ssf's) on the integer lattice by "random wavelet expansion", which stands for representation of random fields by sums of randomly scaled and translated functions, or more generally, by composites of random functionals and deterministic wavelet expansion. To construct ssf's on the integer lattice, random wavelet expansion is applied to the indicator functions of unit cubes at integer sites. We demonstrate how to construct Gaussian, symmetric stable, and Poisson ssf's by random wavelet expansion with mother wavelets having compact support or non-compact support. We also generalize ssf's to stationary random fields which are invariant under independent scaling along different coordinate axes. Finally, we investigate the construction of ssf's by combining wavelet expansion and multiple stochastic integrals.
Year of publication: |
2001
|
---|---|
Authors: | Chi, Zhiyi |
Published in: |
Stochastic Processes and their Applications. - Elsevier, ISSN 0304-4149. - Vol. 91.2001, 1, p. 99-113
|
Publisher: |
Elsevier |
Keywords: | Stationary self-similar Random wavelet expansion Multiple stochastic integral Invariance under independent scaling |
Saved in:
Saved in favorites
Similar items by person
-
On expected occupation time of Brownian bridge
Chi, Zhiyi, (2015)
-
Approximating likelihoods for large spatial data sets
Stein, Michael L., (2004)
- More ...