A Novel Weighting Approach for Variance Component Estimation in GPS/BDS PPP

Abstract

Kalman filter is widely used in precise point positioning (PPP) process to achieve optimal positioning solution. However, the positioning accuracy and stability are closely related to the covariance matrix or weight matrix of observations used in the filtering process. Typically, the weights of pseudorange and carrier-phase types of observations in the PPP are determined based on a priori assumption of noise variance. These imprecise weight values significantly degrade the performance of the filter. In global positioning system (GPS)/BDS PPP, there is an additional challenge to set the weights of GPS and BDS groups of observations since the observation quality of GPS and BDS systems are not the same. In this paper, a novel weighting approach based on the combination of modified variance component estimation (MVCE) and Helmert variance component estimation (HVCE) methods is proposed. The MVCE method is developed to adjust the weights of pseudorange, and carrier-phase types of observations from the same GNSS and HVCE are customized to adjust the weights of GPS and BDS groups of observations. An additional sliding window averaging filter is used in the MVCE to obtain a more precise and stable weight ratio solutions between carrier-phase and pseudorange types of observations. Both static and kinematic tests are conducted to verify the effectiveness and to assess the performance of the proposed weighting approach. The results indicate that the proposed scheme can significantly improve the positioning accuracy and stability as well as reducing the convergence time in GPS/BDS PPP.