Frequency Stability
Mathematical formulas evaluate the frequency stability of precision oscillators and clocks over varying time intervals. The overlap allan variance improves upon the standard variance by utilizing all possible overlapping data segments in the calculation. This statistical tool provides higher confidence and lower uncertainty in the stability estimate.
Statistical Weighting
Data utilization is maximized by sliding the analysis window through the time-series measurements. Instead of jumping by the full stride of the averaging interval, the overlapping method increments the window start by a single sample. This approach yields a significantly larger number of equivalent degrees of freedom from the same dataset.
Calculation Method
A weighted sum of squared frequency differences is computed across the entire measurement record. This calculation incorporates a scaling factor that accounts for the overlapping structure of the data points. Because the samples are statistically dependent, the computation is more intensive but offers a more precise stability curve at longer averaging times.
This precision helps in identifying subtle drift patterns in precision atomic clocks.
Noise Characterization
Error bars are tightened on the resulting stability plots. This reduction in variance makes it easier to distinguish between different noise processes like white frequency noise or random walk frequency noise. It remains standard practice in chronometric validation.
