[time-nuts] Crosscorrelation phase noise measurements
azelio.boriani at screen.it
Thu Aug 2 19:47:06 EDT 2012
Excuse my pop-up into this topic, just let me try to complete: you
cross-correlate and obtain the most probable samples and then do your phase
noise process on those samples. The cross-correlation is only a filter, a
preprocessor for the samples. Am I on the right way?
On Fri, Aug 3, 2012 at 1:31 AM, Magnus Danielson <magnus at rubidium.dyndns.org
> Hi Sylvain,
> On 08/02/2012 11:24 PM, Sylvain Munaut wrote:
>> Hi Magnus,
>> Thanks a lot for the concise and very clean explanation.
> You are welcome!
> The cross-correlation part between the two signal was clear enough in
>> my head but I didn't really see how it would achieve much gain. I
>> didn't think about averaging many resulting spectrum while they're
>> still complex (and not just the amplitudes ... ). I assume that the
>> cross correlation of the two measurement makes the phase of several
>> consecutive measurement "align" so that the main signal accumulates
>> over many averages while the noise is just averaged out.
> Because they correlate, they add up, because the noise does not correlate,
> it flattens out.
> You should look up what is written by NIST on this technique. They have a
> nice online archive.
> They also demoed it on the NIST seminar.
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to
> and follow the instructions there.
More information about the time-nuts