Solving Multi-GNSS Signal Biases from Undifferenced Observations

摘要

The advent of multi-constellation multi-signal navigation satellite systems requires new approaches in the bias analysis of GNSS signals. The state-of-the-art techniques processing dual-frequency GPS and GLONASS data involve taking differences between the signals, which has become unfeasible now that there are half a dozen GNSS systems providing observations over various frequencies and modulations, resulting in over a hundred different observable signals with distinct properties. This work demonstrates joint estimation of all the station and satellite biases along with ionospheric delays, receiver clock offsets and carrier integer ambiguities from undifferenced raw multi-GNSS pseudorange and carrier phase observations. A sparse robust linear least squares solver combined with integer ambiguity resolution was implemented in MATLAB and its feasibility demonstrated using GPS, GLONASS, Beidou, and Galileo observations from MGEX network and post-processed satellite orbit products from ESOC. The results show that the method is computationally feasible and has high enough success rate to justify further study.