Statistical Signal Processing for Atomic Clocks
We have developed signal processing methods to characterize atomic clocks. In particular, we have considered nonstationary behaviors that can arise when an anomaly is taking place. We have proposed a new transform, the dynamic Allan variance, which is a representation of the time-varying stability of an atomic clock. We have shown that the dynamic Allan variance can detect and identify clock anomalies.
The dynamic Allan variance is currently used by the US Naval Observatory to monitor the GPS clocks and by ESA to monitor the clocks onboard GIOVE-A, the first experimental Galileo satellite. It has also been implemented in Stable32, the commercial software standardly used for clock stability analysis.