A atomic time algorithm based on predictable weighting and wavelet multi-scale threshold denoising

The principle of predictable weighting algorithm is that a good clock is a predictable clock, we observed that deterministic signature such as frequency drift and ageing do not affect time scale stability if well predicted in studies of the algorithm. The clock difference is a non-stationary random process, in order to reduce noises of atomic clock, improve the stability of time scale, wavelet muti-scale threshold is studyed to reduce the effects of atomic clock noise, in this paper, the influence of different threshold function for atomic clocks noise reduction is studyed. The advantage of this algorithm is that the predictable weighting is made full use to produce time scale, at the same time, atomic clock noises are reduced, it improve the stability of time scale in many aspects and is an optimiz able algorithm of atomic time scale.