The particle filter based tracking algorithms were researched for the GNSS receiver.Traditional GNSS receivers encounter a difficult problem with high Doppler shift and its derivatives in high dynamic scenarios.To track this shift correctly,the bandwidth of tracking loops must increase,which inevitably induces more broadband noise and may lead to a loss-of-lock at low SNR conditions.In this paper,a novel high dynamic carrier tracking algorithm based on the Gaussian particle filter (GPF) is proposed.The GPF approximates the posterior density by estimating parameters of Gaussian densities based on particle approach at each loop update,which avoids the problem of particle degeneracy.To deal with the problem of divergence and to obtain a better robustness,two optimization methods are also proposed: the STF-GPF and the UGPF.Extensive evaluations on the proposed model and algorithms are conducted with different levels of signal-to-noise ratios (SNRs).The numerical simulation results show that the proposed GPF tracking loop and consistently outperform the conventional approaches,as well as achieving lower SNR sensitivity in comparison with EKF and UKF.Both optimization methods improve tracking accuracy than GPF.