Interpretation of the dynamic Allan variance of nonstationary clock data

Atomic clocks are ultra-stable time references, and they have found fundamental applications in several fields, navigation being certainly one of the most successful. Unfortunately, atomic clocks may experience nonstationary behaviors which make their stability a function of time. Even when the change in stability is thought to be small, it can still force the system that uses the clock to fail the design requirements. In navigation systems this might turn into an unwanted degradation of the user localization error. It is therefore essential to monitor the stability of the clock, and to understand and interpret the possible nonstationary behaviors. To this aim, we have recently Introduced the dynamic Allan variance, or DAVAR, a quantity that allows to represent the evolution with time of the clock stability. In this paper we show several cases of nonstationary time series, made by the combination of random processes and deterministic signals, and we discuss their dynamic Allan variance. The ultimate goal is to be able to do the inverse operation, that is to understand and classify the possible nonstationarities directly from the DAVAR representation. Examples with experimental data are presented.