Simultaneous estimation of inter-frequency clock biases and clock offsets with triple-frequency GPS data: undifferenced and uncombined methodology and impact analysis

Nowadays, multi-frequency signals have been putting forward good prospects for improving the application performances of the rapid-developing Global Navigation Satellite Systems (GNSS). Inter-frequency clock biases (IFCBs), which originate from the discrepancies of both code and phase hardware delays at different frequencies, are therefore important biases that need sufficient consideration and modeling in multi-frequency (triple or more) GNSS data processing. Hence, we have established an undifferenced and uncombined multi-frequency network model, for the simultaneous estimation of IFCBs and clock offsets. In the proposed method, code and phase IFCBs are modeled and re-parameterized as a total IFCB, which is free from external inter-frequency differential code bias corrections and more consistent and convenient for user-side correction. Through the simultaneous estimation of IFCBs and clocks, we can directly study the influences of IFCBs on precise clock offset estimation (PCE). This method overcomes the deficiency of routine geometry-free and ionosphere-free (GFIF) methods, in which clock parameters are eliminated. Additionally, it supports full-channel estimation of IFCBs in a unified procedure based on a network with inhomogeneous receivers. With the proposed IFCB modeling, the accuracy of the estimated satellite clocks could be improved from 42.7 to 22.7 ps. Moreover, after proper code and phase IFCB correction in triple-frequency simulated kinematic PPP, the performances can be significantly improved by at least 29.5% and 21.8%, in terms of positioning accuracy and convergence speed, respectively.