Impact of the anti-aliasing pre-filtering on the measurement of maximum time interval error

The latest ETSI and ITU-T standards recommend the evaluation of the MTIE based on time error data measured, for observation intervals in the range from 0.1 s to 1000 s, “through an equivalent 10 Hz, first-order, low-pass measurement filter”. The impact of such pre-filtering is thoroughly studied by simulating periodical and power-law noise types. Moreover, some results of measurements on telecommunications clocks are provided in order to support the conclusions stemming from simulation with experimental evidence. The results shown point out that the impact that such anti-aliasing pre-filtering can have in practice may be substantial and even misleading: measurements accomplished through such a filter may produce too optimistic results, hiding the true amplitude of the phase fluctuations at the timing interfaces