Exhaustive Statistical Test Package (Version YANT
5.1)
1- This is
the latest and upgraded package from "Statistical Test Library Version
YANT 4.1) it include extra software tools for exhaustive
testing for randomness for multiple samples at a time. The resulting report
will have two dimensional test results, to signify the behavior of different
samples over different tests a conclusive report may be generated to display
the average test results (Pass ratio) for all tested samples. In addition to
the current tests of version YANT 4.1, the Exhaustive Statistical Test Package
YANT 5.1 incorporate the following statistical tests derived from different
statistical institutes:
- The Binary Matrix Rank Test
- Discrete Fourier Transform Test
- Non-Overlapping Template Matching Test
- Overlapping Template Matching Test
- Lempel-Ziv Compression Test
- Approximate Entropy Test
- Cumulative Sums Test
- Random Excursions Test
- Random Excursions Variant Test
- External Devices: None.
The available platforms to run this package are:
-
Any Win9x, WinXP (home edition / professional edition), and Win2K (server / professional).
-
128MB RAM at least.
-
20 MB free HardDisk space.
-
Pentium III.
Operational
Specifications
1-Selection of Test Samples
To select the input test sample(s), the user
may apply either one of the following
methods:
a- Create your own sample file using the
package's Sample Editor. The sample editor enables you to write your test sample either in ASCII
characters format or as bits (1's
and 0's) in binary format.
b- Choose an input file which contains the
output bits under test. Upon selection the file will be loaded to the package.
Besides selecting the input file, the user may select the file format which is
either ASCII characters format or bits in binary format.
c- Select multiple sample files, each file may
represent a different output to be tested. Up to eight sample files may be loaded
at once with the same files' format.
2- Conducting Statistical Tests
The Statistical Tests Library in YANT 5.1
provides many different standard statistical tests. You can choose any of the
statistical tests to be performed on the selected test sample. The tests are
divided into 2 groups based upon whether a certain parameter is required with
the test or not. The 2 groups are:
Non-Parameterized
Tests:
1-
Frequency (Monobit)
Test.
2-
Serial Test.
3-
Poker Test.
4-
Runs Test.
5-
Longest Run of Ones Test.
6-
Binary Matrix Rank Test.
7-
Auto-correlation Test.
8-
Maurer's Universal Test.
9-
Lempel-Ziv Compression Test.
10-
Approximate Entropy Test.
11-
Cumulative Sums Test.
12-
Random Excursions Variant Test.
13-
Discrete Fourier Transform Test.
14-
Random Excursions Test.
15-
Non Overlapping Template Matching Test.
16-
Overlapping Template Matching Test.
Parameterized
Tests:
17-
Block Frequency Test with variable block
size.
18-
Linear Complexity Test with variable
block size.
3- Sample Size Verification:
Before conducting any of the above tests the
application should verify the validity of the acceptable length of the test
sample, each test may imply different sample size for proper operation.
4- Reporting Test Results
For each test the, following results may be appear in the report form:
(1)
Full description of the input test sample
including sample size, file path, sample format and time & date of running
the tests.
(2)
The important calculated values within
the test.
(3)
The calculated test statistic as well as
the threshold (tabulated) value from which the user can evaluate the test
result and determine whether the test passed or failed.
(4)
The final test result (either passed or
failed).
(5)
A graphical representation of the
chi-square distribution of the computed and tabulated values along with the
threshold value.
5-
Saving/Loading and Printing Functions
The final report for each conducted test may
be saved for further referencing and viewing. The file extension for the
Statistical Tests Reports may have the extension (.str).
The user may have the option to load any previously saved Statistical Test
Report at any time . The printing option as well is
valid for any report.
See Online Demo
|