Efficient frequency jumps detection algorithm for atomic clock comparisons

In this paper a new method of frequency jumps detection in data from atomic clock comparisons is proposed. The presented approach is based on histogram analysis for different time intervals averaging phasetime data recorded over a certain period of time. Our method allows identification of multiple frequency jumps for long data series as well to almost real-time jump detection in combination with advanced filtering. Several methods of preliminary data processing have been tested (simple averaging, moving average and Vondrak filtration), to achieve flexibility in adjusting the algorithm parameters for current needs which is the key to its use in determining ensemble time scale or to control systems of physical time scales, such as UTC(PL). The best results have been achieved with the Vondrak filter.

---

1. This work was supported by the Polish Ministry of Science and Higher Education under the project of Special Research Equipment support “Assembly of Time and Frequency Standards for participation in the Polish Atomic Timescale TA(PL)” (dec. no. 8/E-242/SPUB/SN/2019 and 6/E-242/SPUB/SN/2020).

---

2. Opracowanie rekordu ze środków MNiSW, umowa Nr 461252 w ramach programu "Społeczna odpowiedzialność nauki" - moduł: Popularyzacja nauki i promocja sportu (2021).