Research On The Methods Of GNSS Data Processing Based On Distributed Computing Technology

With the rapid development of satellite navigation and computer communication technology, the basic theory and application of GNSS data processing are in the process of further diversification and improvement. At the same time, distributed computing provides a new method for current massive data processing, which is highly efficient. Detailed researches and practice are carried out in this paper, the main work and conclusions are summarized as follows:1. The basic theory of GNSS data processing, the application of data processing software and the background and current status of distributed technology in the geodesy are analyzed, the main problems and solution are pointed out, and the scope and basic content of this paper are defined.2. The content and the characteristics of the theory of the GNSS observation data processing are introduced. As the basic models in GNSS precise positioning, the Precise Point Positioning and the double difference relative positioning are explained and analyzed. The concept of distributed computing is introduced, and the distributed key technology of interoperability, virtualization and Web Service are displayed, which is the foundation of its application in GNSS data processing.3. The improved pretreatment and distributed applications of the measurement data preprocessing are researched. On the basis of introducing the theory and disadvantages of TurboEdit in cycle slip detection and repair, the algorithm is improved by optimizing the detection method and good results have been achieved. The function and usage of TEQC software are analyzed and its visualization is developed, and a new TEQC automation preprocessing mode based on the time window is put forward. In the environment of distributed network, the high efficiency of quality checking and preprocessing operation is verified under the improved data preprocessing module.4. The distributed computing technology is applied to GNSS network data processing. The common data processing efficiency of un-differenced PPP and double differenced network solution is analyzed. Under the environment of distributed network, the distributed processing of un-differenced PPP and the sub-network division of GNSS reference station network are achieved by calling current GNSS software or developed processing model. The new network solution algorithm of Ambizap is detailed derived. On the basis of distributed processing of PPP and baselines’ double difference ambiguity solutions, the PPP solutions can be constrained by double difference fixed solutions, and the whole network adjustment can be achieved. In this way, the accuracy can be assured and the computational efficiency can be effectively improved.5. Based on the theory of GNSS data processing and distributed computing technology, combined with centralized application of data processing software, the GNSS surveying data processing software prototype is designed and developed. And the function of data management, data preprocessing and data positioning solution can be achieved under the distributed interoperability system, which is convenient for the sharing and interoperability of GNSS sources.