Enhanced stability for local atomic clock ensemble time scale using weighted moving average filter

This paper presents a method to enhance the stability of atomic clock ensemble time scale by using weighted moving average filter. Compared with Kalman filter, the weighted moving average filter has simpler algorithm structure, higher computational efficiency. Its convergence time is shorter than Kalman filter. When the two filters were applied to frequency data of Cs clocks, they have consistent filtering performance. The preliminary results show the weighted moving average filter can enhance the stability of local atomic clock ensemble time scale, especially the short- and middle-term stability.