Abstract: Time comparison realized by precise point positioning (PPP) can be easily influenced by random noise. Therefore, it is essential to separate noise from signals in time series. In this paper, an empirical mode decomposition (EMD) based de-nosing method for time comparison data firstly is presented. Furthermore, the proposed method is compared with the Vondrak filter. It is shown that the developed method is comparable with the Vondrak filter, and can improve the stability of time transfer noticeably.

Key words: Time Comparison, Precise Point Positioning (PPP), Empirical Mode Decomposition (EMD), De-nosing