Parallel Computation of Multi-GNSS and Multi-Frequency Inter-Frequency Clock Biases and Observable-Specific Biases

With the widespread application of GNSS, delicate handling biases among different systems and frequencies is critical importance, wherein inter-frequency clock (IFCBs) observable-specific signal (OSBs) should be carefully corrected. Usually, a serial approach used to calculate these products. To accelerate computation speed reduce time delay, multicore parallel estimation strategy for IFCBs, code, phase OSBs by utilizing task library (TPL) proposed, computations, including precise point positioning (PPP), estimation, being carried out on basis data parallelisms task-based asynchronous programming. Three weeks observables from multi-GNSS experiment campaign (MGEX) network utilized. The result shows that IFCB errors GPS Block IIF GLONASS M+ satellites are nonnegligible, in which satellite R21 largest more than 0.60 m, while those other marginal, code present excellent stability with standard deviation (STD) 0.10 ns approximately 0.20 systems. Besides, all show better ns, Galileo best performance 0.01 ns. Compared single-core computing method, acceleration rates IFCBs 3.10, 5.53, 9.66, 17.04 times higher using four, eight, sixteen, thirty-two physical cores, respectively, through multi-core parallelized execution.