Validating frequency transfer via interferometric fiber links for optical clock comparisons

We investigate the validation of fiber-based optical frequency transfer for frequency comparison applications. We specifically consider the frequency transfer validation for remote optical clock comparisons and want to ensure interferometric fiber link uncertainty contributions below the combined uncertainty of the clocks under test. The validation is based on signals obtained via looping back from the remote end to the sender site and comparing the input with the output of the loop. These loop-back data need to be averaged over intervals for reaching the validation goal, as the short-term instability of long-distance interferometric fiber links is typically higher than that of optical clocks. We introduce a two-step validation approach and address the finding of a compromise between opposing aspects of averaging: reaching low uncertainties versus achieving a high data coverage of the validated data set via a high temporal resolution of the fault identification. We discuss the impact of different averaging types and of the tolerance of filtering criteria on the achievable estimated uncertainty and on the coverage of the validated data set. Data from four multiple-week-long measurement campaigns on the fiber link between Physikalisch-Technische Bundesansanstalt and University of Strasbourg are used for this assessment.