Research On GNSS Baseband Technology Of Scalar Deep Integration For High Accuracy Positioning In Urban Harsh Environment

With the blooming of navigation and location based service(LBS)industry and the fast growing automobile market in China,the requirement of vehicle navigation in urban harsh environment keeps increasing.Currently,automatic driving and unmanned aerial vehicle(UAV)also have urgent demand on continuous,reliable decimeter even centimeter level navigation.Global Navigation Satellite System(GNSS)is one of the crucial methods.The industry is actively developing high precision,low cost GNSS positioning,navigation and timing terminal products.Traditional GNSS navigation receivers focus on high sensitivity in urban harsh environment and ignore the precision of GNSS observations;while surveying receivers can only be used in high precision positioning in open-sky environment.Based on inertial navigation system(INS)and GNSS,integrated navigation is widely used in land vehicle navigation.However,traditional integration methods are loose integration and tight integration,which are both integration methods on the level of data processing and mainly use GNSS to aid INS.They have no improvement to GNSS receiver baseband signal processing.In order to achieve high precision navigation in urban harsh environment through a low cost method,the quality of GNSS raw measurements should be increased.Deep integration uses INS information to aid receiver baseband signal processing.Therefore it can effectively improve the dynamic response and the tracking precision of receiver baseband signal processing at the same time.Because of its inherent superior dynamic performance,the previous research and application on GNSS/INS deep integration system focuses on high dynamic scenarios and strong jamming environment,and it has been applied to smart weapon guidance by US and NATO.With the extension of the deep integration technology to civilian applications and the development of land vehicle navigation,some foreign manufacturers have developed high precision navigation systems with quasi deep integration technology.Towards the high precision and low cost land vehicle navigation in urban environment,the thesis is supported by the National High-tech R&D Program(863 Program)project"Development and applications of land vehicle navigation terminals based on GNSS/INS deep integration".Aiming at intermittent and attenuated GNSS signals in urban harsh environment,this thesis studies the theory and methodology of inertial aided GPS and Beidou receiver using GNSS/INS deep integration;extracts continuous high precision GNSS measurements in urban vehicle navigation;and evaluates the measurements quality of the GNSS/INS deep integration receiver through both hardware simulator and field tests.The research includes the following aspects:1.A unified and comprehensive baseband tracking error model for GNSS/INS deep integration is proposed.The inertial Doppler frequency aiding error model is derived by substituting the inertial sensor error model into the differential equation of INS error.Then the overall error response of the deep integration tracking channel is derived by substituting this inertial Doppler frequency aiding error model into the receiver tracking error model of close-loop and open-loop.A statistical characteristic expression of the error response of the items modeled as stochastic processes in inertial sensor error model is presented.By doing the above,a theoretical tool for the quantitative analysis of close-loop tracking error and open loop tracking time limit is provided.2.The weak signal tracking and dynamic tracking performance of the FFT discriminator is studied.The IQ integration sequence without navigation data bits is used to perform the FFT transformation for the case of having navigation data bits aiding information.In the absence of navigation data bits aiding information,navigation data bits of the IQ integration sequence are removed by complex squaring.The FFT transformation is then performed and the generalized non-coherent PLL carrier phase discriminator is derived.The weak signal sensitivity is improved by using partial frequency spectrum points FFT,and a low complexity implementation of the partial FFT discriminator which is suitable for hardware is proposed.By using the probability density function of the IQ integration with and without GNSS signal,the sensitivity of the FFT discriminator is estimated.By using Monte-Carlo simulation,the sensitivity of the squared FFT discriminator is simulated.By combining the acceleration estimation error of inertial sensor with the FFT discriminator dynamic performance,the dynamic sensitivity performance of the FFT frequency discriminator in deep integration is analyzed.Hardware simulation results have shown that reliable GNSS signal tracking at 20dB-Hz weak signal tracking in 100g acceleration/deceleration scenario is achieved.3.The GNSS observation continuity in satellite signal discontinuous environment is improved effectively by using inertial aided open loop tracking algorithm.Hardware simulator tests verified the possibility of short-term(5 s)carrier phase open loop tracking.Inter-channel aiding increased the carrier phase open loop tracking accuracy.Hardware simulator tests and urban harsh environment field tests verified the possibility of relatively long-term(20s)carrier frequency and code phase open loop tracking.Reacquisition is accelerated significantly.The total tracking percentage of GNSS signal locking time increased.Code phase tracking error during convergence decreased and the navigation positioning availability and precision is improved.4.Towards the continuous high precision observation requirement in urban harsh environment,a GNSS/INS deep integration software defined receiver is developed and the system is optimized.In addition to the above key techniques,the following designs and improvements are included:1)Zero padding FFT is used as the fast acquisition algorithm to avoid being affected by the NH code in Beidou signal when implementing Beidou signal processing.2)The ratio of code frequency and carrier frequency is used as the frequency lock detector.By combining the code lock detector and inertial aiding information with this ratio,a robust signal tracking lock detector is implemented.3)The coherent and non-coherent integration energy of different bit phases is used to synchronize the navigation data bit phase.Further coherent integration is performed by performing FFT transformation on the expanded frame synchronization head to realize reliable frame synchronization in weak signal.4)The code tracking loop with adjustable loop bandwidth is used to accelerate the code phase tracking error convergence.By using code phase step response error,the bandwidth changing time is calculated.5)Pseudorange multipath error is suppressed by using inertial aided narrow correlator.6)By differencing the inertial derived pseudorange and channel extracted pseudorange,a low complexity pseudorange fault detection algorithm is implemented.A low pass filter is used to obtain the slowly varying component of satellite position calculation error,and the residual ionosphere,troposphere delay,receiver and satellite clock bias error.5.The developed GNSS/INS deep integration software defined receiver is tested and evaluated,including simulator signal tests and urban environment field tests.The baseband code and carrier phase tracking performance under different loop settings is compared by hardware simulator signal.The pseudorange and carrier phase measurements are compared with typical commercial receivers(Trimble R9 surveying receiver and ublox M8N navigation receiver).The simulation tests show that the deep integration receiver has higher carrier phase tracking sensitivity,smaller carrier phase tracking error,and shorter reacquisition time after the GNSS signal recovery,compared to R9 receiver.The deep integration receiver also has similar pseudorange error,similar reacquisition time,and smaller code error during convergence,compared to ublox receiver.Real urban harsh environment signal test is also performed.Firstly,positioning results of the GNSS/INS deep integration receiver working in inertial aided mode and unaided mode are compared.With the aiding,the deep integration receiver developed in this thesis has a better performance in horizontal positioning error,fix epoch percentage and coarse error epoch percentage.Secondly,pseudorange positioning results of the inertial aided deep integration receiver developed in this thesis are compared with typical commercial receivers.Deep integration receiver has a better performance than Trimble R9 and ublox receiver or is similar to them in terms of horizontal positioning error,fix epoch percentage and coarse error epoch percentage.Then,pseudorange positioning results in some typical cases are compared,and the inertial aided deep integration receiver developed in this thesis has a better performance than R9 and ublox receivers or is similar to them in terms of convergence time,fix epoch number and horizontal positioning error.Lastly,carrier phase positioning results in a relative harsh environment are compared,and the deep integration receiver has a better performance than R9 and ublox receiver in terms of horizontal positioning error and ambiguity fix number.In summary,the scalar GNSS/INS deep integration techniques dedicated for urban harsh environment is developed and studied thoroughly in this thesis.The deep integration receiver baseband error model is improved and become more comprehensive and feasible.The dynamic weak signal tracking performance of the FFT frequency discriminator is analyzed.The open loop tracking based on inertial aiding is implemented to improve the continuity of the GNSS signal tracking.The deep integration receiver is optimized for harsh environment and complete hardware simulation and urban harsh environment land vehicle tests are performed.The deep integration receiver developed in this thesis has low computational complexity,which can be ported easily to embedded system,and can work as the validation tool and test platform for the development of real-time hardware receiver products.The research works done in this thesis has improved the GNSS receiver signal processing capability in harsh environment of land vehicle dynamic scenario significantly.They have provided a high quality GNSS observation level that can guarantee the continuous high precision positioning in urban environment,demanded by the fast growing driverless car and UAV industry.