Continuous Smoothing Positioning Algorithm Based On GNSS Observations Of Smartphones

At present,the Global Navigation Satellite System(GNSS)chipset is embedded in almost every smart terminal such as smartphone and tablet.Based on its derived Location Based Services(LBS)application,such as sharing location in social software,pedestrian/car navigation,precision marketing,etc.,enrich everyone's daily life more and more.In May 2016,Google company announced that the developers could access some Android GNSS raw measurements,including the pseudorange,carrier phase measurements and navigation massages in Android 7.0.It means that the GNSS chipset embedded in terminal is no longer a “black box” for developers,which only provide position,speed and direction.Therefore,developers could make deeper optimizations on the navigation and positioning performance of smart terminals.In addition,the quality of the GNSS hardware for smartphones is constantly improving.Since 2018,manufacturers have lauched dual-frequency L1/L5 smartphones,such as Xiaomi 8,Huawei MATE20 and so on.The addition of L5 frequency observations enable achieve dual-frequency L1/L5 ionospheric free combination to eliminate ionospheric delay effects.In addition,L5 frequency observation has the characteristics of longer wavelength,smaller signal intensity attenuation,longer and faster symbol period,etc.It would be more resistant to multipath effects and more suitable for urban challenge environment navigation and positioning.However,unlike GNSS surveying receivers,smartphones typically use small linear polarized antennas and low power GNSS chipsets.It would definitely affect the quality and characteristics of GNSS observation,which would affect the accuracy and reliability of the solution.This paper conducts a comprehensive and in-depth quality analysis of GNSS observations of single/dual-frequency smartphones commonly found on the market.Then,combined with its unique quality characteristics,a continuous smoothing positioning algorithm based on GNSS observations of smartphones is proposed.The feasibility of the algorithm is verified by the dynamic data of related single/dual frequency smartphones.The main work and conclusions of this paper are as follows:(1)This paper introduces the research background the navigation and positioning of mass-market smartphone;summarizes the difficulties of restricting the precise positioning of smartphones;summarizes the current research status about quality assessment and positioning algorithms on smartphone raw GNSS observations in recent years.(2)Introduce the space-time reference of GPS,BDS,GALILEO,GLONASS,QZSS systems and give a unified method;for the single and dual frequency mass-market smartphones,“L1/L5 ionosphere-free combination” and "L1/L5 un-combination",basic models of multi-frequency and multi-system standard point positioning,are given;Introduce the basic model of Kalman filtering for parameter estimation.(3)Introduce the basic structure of the smartphone's GNSS module and the acquisition and generation method of Android raw GNSS observations;The comprehensive indepth analysis of single/dual frequency is done.The assessment mainly includes the satellite tracking,signal-to-noise ratio,pseudorange single-differenced residual,duty cycle,and velocity estimation,etc;According to the analysis,the accuracy of GNSS pseudorange observation of the smartphone is lower and easy to be gross error than surveying receiver.In addition,the accuracies of the phase and Doppler are relatively precise,and its accuracy of the velocity estimation could reach cm/s in some cases.(4)According to the quality characteristics of pseudo-range observations unique to smartphones,the corresponding stochastic models are established,and some data preprocessing strategies for the original GNSS observations of smartphones are given.(5)Based on the analysis of the raw GNSS observations of smartphones,a continuous smoothing positioning algorithm for single/dual-frequency smartphones considering the duty cycle based on GNSS raw observations is proposed.The essence of the algorithm is to use accurate carrier position changes,provided by time-differenced from carrier phase or doppler velocity,smooth the noise of the pseudorange in standard point positioning.Subsequently,the corresponding algorithm verification was carried out by using the dynamic experiments of the single-frequency Google Nexus 9,Huawei P10,Samsung S8,dual-frequency Huawei MATE20,and Xiaomi 8 smartphones.The results show that the single-frequency positioning accuracy can reach 3-4m in the plane and 3-5m in the elevation direction.The positioning accuracy of the dual-frequency mobile phone is better than that of the single frequency,and the accuracy of the plane and elevation are 2-3m and 3m respectively.The positioning result of the singlefrequency/dual-frequency smartphones is continuous,smooth,and could guarantee good navigation user experience.