Research On The Key Technologies Of Ionosphere Integrity Of BeiDou Satellite Based Augmentation System

For aviation and other Safety-of-Lift users,the integrity of navigation system is a requirement that must be considered.As the ‘patch' of the traditional GNSS,the Satellite Based Augmentation System(SBAS)can provide augmentations on accuracy and integrity.The modeling and estimation of ionosphere characteristics has always been the focus of research in the field of SBAS.The existing SBAS ionosphere model and enhancement parameter estimation are based on the analysis of data from CONUS region.The directly use of the model to the BD SBAS may lead to integrity risk and the decrease of correction accuracy.In this context,this paper focuses on the estimation of Grid Ionosphere Vertical Error(GIVE),which is a parameter related to the ionosphere integrity,in order to improve the availability of the system while satisfying the integrity requirement.The main contents and innovations of this paper include:Firstly,the applicability of the traditional SBAS ionosphere model in the service area of BD SBAS is analyzed.This paper focuses on three aspects,namely,the deterministic part,the random part and the system's time and space resolution requirements.Results are compared with that in the US mainland region.The results show that in terms of functional form,the linear planar model is still established in the BD SBAS service region,but the fitted residuals are larger than those in the continental US,which is related to the characteristics of the ionosphere in the geomagnetic low-latitude region.In terms of the statistical characteristics of the random part of the model,the model noise in BD SBAS area has a larger noise variance and exhibits a more obvious non-Gaussian character.In terms of space and time resolution,although reducing the grid interval and time update period can improve the accuracy of the model,the performance improvement is not obvious relative to the complexity of the SBAS interface modification,so these two parameters are maintained in the analysis in this paper.Secondly,this paper proposes a method based on linear planar model with bias detection to solve the problem that the estimation error of the traditional minimum variance search DCB estimation algorithm is large at low geomagnetic latitude and ionosphere active period,and the advantages of this method are proved theoretically.The IGS data was used to analysis the DCB estimation performance of the proposed method in different geomagnetic latitudes,different seasons,and different ionospheric activities.The results are compared with the traditional method.The results show that there is no obvious seasonal difference in DCB estimated performance.Geomagnetic latitude and ionospheric activity are the key factors affecting the performance of the traditional minimum variance search algorithm.Taking the analysis result of the 30-day DCB estimation stability of BOGT station at low latitude as an example,the standard deviation of DCB fluctuation in this algorithm is 1.49 ns,which is much smaller than the standard deviation of 5 ns in traditional methods.At the same time,the analysis results of 100 IGS stations around the world show that under the active ionosphere status,the average deviation between the DCB estimation results of the proposed algorithm and the products distributed by the European Orbiting Center(CODE)is 1ns,which is less than the 2ns under the traditional method.Thirdly,in order to solve the problem that the estimation results of undersampled error variance based on Relative Centroid Metric(RCM)in WAAS are too conservative,this paper concludes the characteristics of a ‘good' metric from the physical point of view,and proposes an ionosphere threat model based on Relative COVerage(RCOV)metric.The design parameters of the threat model are given theoretically,and the experimental results are carried out.The results show that during the quiet period of the ionosphere,the RCOV model can provide more than 20% usability performance improvement over RCM at about 36% of the effective area.During the active period of the ionosphere,the performance of usability is improved even more.At the same time,by analyzing the statistical characteristics of the normalized ionospheric delay estimation error,the article has initially verified the integrity of the system.Fourthly,a Modified Paired Overbound(MPO)method is proposed to solve the problem that the existing overbound methods are too conservative and lead to the lack of system availability.At the same time,considering the zero-mean Gaussian hypothesis in GIVE estimation and protection level calculation in the current SBAS MOPS,this paper deduces the method for calculating Gaussian distribution parameter under zero-mean constraint with PO method.Through experiments,the performance difference between the traditional method and the improved method is compared.The results show that in the measurement domain,the error model obtained by the MPO method has smaller differences between the empirical distributions,and the performance improvement is about 50% compared to the traditional method.The amplitude of improvement will increase with the addition of convolution times.The difference between the overbound results and the empirical distribution is also related to the empirical distribution,and the closer the empirical distribution of the sample data is to the Gaussian distribution,the smaller the difference is.