Gnss Original Data Compression Method Based On State And Residual Representation

In the Global Navigation Satellite System,satellites transmit data which is preserved as Receiver Independent Exchange Format to Continue Operation Reference System.At present,most of the GNSS files store and transmit the original observation data directly,but they themselves have very high redundancy.In the Satellite systems trend of the distributed,globalization,diversified,all-weather,high sampling and multi band,the observation data are massive outbreaks,which bring great challenge to the processing,storage and transmission of the data.Currently,there are not many methods to compress GNSS observation data.Most of them use recursive difference method to compress data according to the time correlation between epochs in the data.Based on the connotation of data space,this paper proposes a new compression method.Using the kalman filtering method,the observation vector is mapped from the observation space to the state space to reduce the data dimension,and the residual vector is generated,that is,the process of un-differenced point positioning.Then,the differential compression method is used to weaken the temporal correlation of the state and residual vectors,so that the amount of data is significantly reduced compared with the observation vector.This compression method not only solves the shortcomings of the traditional compression method which must be decompressed and reduced the original observation file,but also breaks the low compression rate of the compressive sensing theory.In this paper,we use Melbourne-Wubbena combination method and Geometry-free combination method to detect cycle slips.After the data preprocessing is completed,the error correction models are corrected according to the relevant error correction models released by IERS2010.The Kalman filter is used to estimate the state parameters.Convergent results are obtained.Then we calculated the internal coincidence accuracy in the three directions of the station coordinates.The X direction is 0.235 m,the Y direction is 0.276 m,and the Z direction is 0.103 m.In addition,the residuals of pseudorange range from-5m to 5m,and the residuals of phase is between-5cm and 5cm.After compressing the original observation values of GNSS,a set of special format for GNSS original data compression based on residual and state parameters is designed.This format supports each error,single or multiple transmission of residual data.If the compression format can be used directly at the sensor,it will greatly reduce the energy consumption of data transmission.Through the double difference theory,it is proved that the data can be processed directly from the residual and state parameters in the compression result,but it is not necessary to restore the original file through the decompression process,and then the calculation result is independent of the positioning accuracy.At the same time,the compression method conceals the coordinates of the survey stations,and the users can not solve the station coordinates according to the state parameters and residuals,thus promoting the confidentiality of the station coordinates.Finally,compared with the crx2rnx/rnx2 crx software,it is concluded that the compression method proposed in this paper has a higher compression rate,about 17.5% of the original observation file,which achieves the purpose of more efficient data compression.