Kalman filter simulation and characterization of BDS satellite clock

Abstract

In this paper, a Kalman filter algorithm for simulation and characterization of BDS (beidou satellite system) satellite clock in orbit is presented. The algorithm is designed to produce near optimal estimation of characteristic parameters for BDS satellite clock. Besides normal clock noises, white frequency modulation, random walk frequency modulation and random run frequency modulation, the flicker phase modulation and flicker frequency modulation are also considered as the linear combination of several Gauss Markov processes. Additional states are used to model the periodic variations of BDS satellite clock, and the first type model of clock periodic states is used in the paper. It is shown that simulation results fit the real oscillator behavior well.