Independent Randomness Tests based on the Orthogonalized Non-overlapping Template Matching Test
Statistics Theory
2019-08-21 v1 Statistics Theory
Abstract
In general, randomness tests included in a test suite are not independent of each other. This renders it difficult to fix a rational criterion through the whole test suite with an explicit significance level. In this paper, we focus on the Non-overlapping Template Matching Test, which is a randomness test included in the NIST statistical test suite. The test uses a parameter called "template" and we can consider a test item for each template. We investigate dependency between two test items by deriving the joint probability density function of the two p-values and propose a transformation to make multi test items independent of each other.
Cite
@article{arxiv.1908.07145,
title = {Independent Randomness Tests based on the Orthogonalized Non-overlapping Template Matching Test},
author = {Atsushi Iwasaki},
journal= {arXiv preprint arXiv:1908.07145},
year = {2019}
}