Enhanced predictability of hydrogen maser using random pursuit strategy

The predictability of atomic clock is generally used to characterize its overall performance. Anomalous clock behaviors lead to greater prediction error in traditional prediction method, such as polynomial fitting predictor and Kalman filter predictor. Our previous work developed random pursuit strategy (RPS) to avoid or minimize above the problem. It was further applied to real data of three Hydrogen masers to verify its effectiveness. Compared to Kalman filter predictor worked on suitable condition, the RPS showed a certain advantage, particularly for clock data with phase or frequency jumps. It could be helpful for steering and weighting algorithms in time keeping.