Developer Reference

Contents

Two-Level Testing

To test
K
subsequences
u
1
,
u
2
, ...,
u
n
;
u
n+1
,
u
n+2
, ...,
u
2
n
; ...;
u
(K-1)n+1
,
u
(K-1)n+2
, ...,
u
Kn
of the original sequence, p-values
p
1
,
p
2
, ...,
p
K
are computed. Since the resulting p-values for the sequence
u
1
,
u
2
, ...,
u
Kn
of truly random numbers are supposed to be uniformly distributed over the interval  (0, 1), a uniformity test can be performed for these p-values, returning p-value
q
1
of the second level. Repeating this procedure
L
times results in obtaining
L
p-values of the second level
q
1
,
q
2
, ... ,
q
L
on which you can perform threshold testing.
We have conducted threshold second level testing for the VS generators with ten iterations (
L
=10) and applied the Kolmogorov-Smirnov goodness-of-fit test with Anderson-Darling statistics to evaluate
p
1
,
p
2
, ...,
p
K
uniformity.

Product and Performance Information

1

Performance varies by use, configuration and other factors. Learn more at www.Intel.com/PerformanceIndex.