23 reputation
3
bio website
location
age
visits member for 7 months
seen Mar 21 at 22:35

Mar
19
comment Randomness test question from FIPS 140-1 and comparison with 140-2
Some good points thanks. I was thinking to use the 140-1 thresholds for testing the raw source output and the tighter 140-2 thresholds for testing the data after Von Nuemann extraction. My system will only allow output of the uniformly distributed extracted data after it's successfully passed the 140-2 tests. Testing the raw source seems more of an academic exercise/sanity test to make sure it's not outputting complete rubbish. If the extracted data passes all the 140-2 tests I think that's perfectly fine for use even if the raw source data failed the tests and wasn't uniformly distributed.
Mar
18
awarded  Scholar
Mar
18
accepted Randomness test question from FIPS 140-1 and comparison with 140-2
Mar
18
comment Randomness test question from FIPS 140-1 and comparison with 140-2
Thanks for your answer. I think adding some of those NIST tests might be useful as well.
Mar
18
comment Randomness test question from FIPS 140-1 and comparison with 140-2
Thanks @poncho and fgrieu. I am testing a TRNG source so I wanted some rough tests to gauge its performance. I was originally testing every 20,000 bits of the raw 12 million bit output using FIPS 140-1 tests. That passed all tests no problem. After switching to FIPS 140-2 it failed only by a very small amount in the Runs Test within 1-2 of the 20,000 bit groups. After running it through the Von Nuemann extractor it seems to pass all the tests just fine. I think if I just use the extracted output it will be fine for my purposes.
Mar
18
awarded  Supporter
Mar
17
awarded  Student
Mar
17
asked Randomness test question from FIPS 140-1 and comparison with 140-2