Global navigation satellite system real-time processing requires low latency, high timeliness, and high computational efficiency. A typical application is providing corrections using data from a regional Continuously Operating Reference Station (CORS) network. Usually the wide-lane and narrow-lane fractional cycle biases (FCBs) are determined at the server and broadcast to users to fix undifferenced ambiguity. Also, a tropospheric model is established at the server and broadcast to users to obtain accurate and reliable a priori zenith total delays for precise point positioning (PPP) using the ionospheric-free (IF) observation combination. Currently, serial methods are typically applied, i.e., all reference stations are involved in estimating the wide-lane and narrow-lane FCBs and establishing a regional tropospheric delay model. To improve the efficiency and shorten the latency, we develop a parallel computation method for regional CORS network corrections based on IF PPP by adopting a multicore parallel computing technology task parallel library, wherein parallel computations involving the FCBs, tropospheric delays, and tropospheric model are successively performed based on data parallelism, in which the same operation is performed concurrently on elements in an array, and task parallelism, which refers to one or more independent tasks running concurrently. Data covering four seasons from the Hong Kong and southwestern America CORS networks are utilized in the experiment. The single differenced FCBs between satellites are determined within each full pass, and a tropospheric model with an internal accuracy better than 1.4 cm and an external accuracy better than 1.6 cm is derived at the server. With the parallel implementation, the speedup ratios of FCB estimation and tropospheric modeling are 1.79, 3.15, 5.59, and 9.69 times higher for dual-core, quad-core, octa-core, and hexadeca-core platforms, respectively, than for a single-core platform.
This is a preview of subscription content, log in to check access.
The authors thank the Hong Kong Surveying and Mapping office and National Geodetic Survey for offering GNSS data and IGS for offering precise products. This study was supported by the National Natural Science Foundation of China (No. 41674019) and the National Key Research and Development Program of China (No. 2016YFB0501701).
Brockmann E, Troller M (2002) GPS meteorology in the Swiss Alps: interpolation accuracy for different Alpine areas and near real-time results. Paper presented at Exploitation of Ground-Based GPS for Meteorology, COST Action 716 Workshop, GeoForschungsZentrum Potsdam, Potsdam, Germany, 28–29 JanGoogle Scholar
Ge M, Zou X, Dick G, Jiang W, Wickert J, Liu J (2010) An alternative network RTK approach based on undifferenced observation corrections. In: Proceedings of the ION GNSS 2010, Portland, Oral reportGoogle Scholar
Serpelloni E, Casila G, Galvani A, Anzidei M (2006) Data analysis of permanent GPS networks in Italy and surrounding region: application of a distributed processing approach. Ann Geophys 49(4/5):897–928Google Scholar
Villiger A, Dach R (eds) (2018) International GNSS service technical report 2017 (IGS annual report). IGS Central Bureau and University of Bern; Bern Open Publishing. https://doi.org/10.7892/boris.116377
Wübbena G, Schmitz M, Bagge A (2005) PPP-RTK: precise point positioning using state-space representation in RTK networks. In: Proceedings of the ION GNSS 2005, Long Beach, California, USA, September 13–16Google Scholar
Yang Y (1993) Robust estimation theory and its applications. Bayi Publishing House, BeijingGoogle Scholar