Research On The Tracking Algorithm Of High Dynamic And Weak GPS Signal

GPS (Global Positioning System), which is also known as GlobalSatellite Positioning System, is the most perfect satellite navigationsystem in the world. Currently GPS provides an accurate and continuousthree-dimensional coordinate information and speed information as wellas timing service to GPS users on as much as98%of the Earth’ssurface. Nowadays, the use of the navigation devices based on GPS isbecoming more and more widespread and has potential for futuredevelopment success. The tracking process, which is one of the mostimportant parts of GPS signal processing, is a real-time tracking of codephase changes and the carrier Doppler changes based on the signalacquisition process, enabling the receiver’s local signal to the exactsynchronization with the input signal. The conventional GPS signaltracking loop using DLL(Delay Locked Loop) and PLL (Phase LockedLoop) to achieve the tracking of code and carrier phase. However, underhigh dynamic or weak signal conditions, the performance of theconventional tracking loop is greatly decreased due to its intrinsic defect.First, under high dynamic condition, conventional GPS signaltracking method has a high probability of losing lock as well as a low accuracy as a result of the big Doppler shift. A new tracking model wasproposed by using the output from a large number of correlators whichspan wide ranges of uncertainty in code phase and carrier phase. Anoptimized iterated Extended Kalman filter algorithm based on theLevenberg-Marquardt method was also implemented here to process theoutput from the correlators, despite large acquisition errors. Thesimulation results show that the new model performances well fortracking a high dynamic GPS signal with acceleration of150g.Also, conventional tracking method is difficult to track weak GPSsignals and has a high probability of losing lock. In order to solve theproblem, a new tracking method was proposed. This work used adaptiveKalman filter to reduce the impact of noise, also, Levenberg-Marquardtmethod was implemented to improve stability and convergence rate of theadaptive Kalman filter algorithm. The simulation results show that thenew method can maintain lock on a signal as weak as21dB-Hz with hightracking accuracy and sensitivity.This paper firstly made a detailed analysis of GPS signal trackingalgorithm. Improved model and algorithm were proposed for each of thetwo different application environments and feasibility of the algorithmwas verified by experiments. It will lay a solid foundation for theapplications of GPS signal for more complex cases in the future.