Decoupled-Clock Model (DCM) improves the accuracy of Precise Point Positioning (PPP) by enabling Ambiguity Resolution (AR). Clock and bias corrections are typically estimated at 30-seconds rate to reduce the computation time. However, high rate PPP-AR applications, typically use 1 Hz tracking data, and ideally require 1-s clock corrections. The quality of the interpolated DCM corrections are evaluated by comparing the interpolation error of DCM clocks and biases of GPS satellites over different intervals. It is showed that for short-term intervals i.e. from 5 to 30 seconds, the RMS error of DCM clock grows by 0.5, 2.5, 2.8, 3.2 and 5.6 ps per 5-second increments of the interpolation interval respectively for Blocks IIF-Rb, IIA, IIR, IIR-M and IIF-Cs. The RMS interpolation error of DCM biases were dominated by pseudo-range noise where the noise level of ionosphere-free code/phase biases were about ～4.2 times of the noise level of wide lane biases. Allan Deviation (ADEV) of DCM corrections are then estimated to evaluate the short-term stability and to characterize the random behaviour of these corrections within different intervals. This allows the high-rate PPP-AR users to introduce observation weights based on the expected clock instabilities and interpolation errors. The deterioration of the horizontal PPP-AR on kinematic mode are compared by applying DCM corrections at the rate of 5, 10...30 seconds. In kinematic PPP-AR, at least 60 minutes of GPS observations were required to reach 80% AR success-rate that allow for centimeter-level horizontal precisions. Using 5- and 30- second DCM corrections increased the RMS error of PPP-AR by respectively 1, and 6 mm that represents 10% and 60% deterioration comparing to 1-second solutions.
展开▼
