Research And Simulation Of Baseband Signal Processing Of High-Sensitivity GPS Receivers

With the spreading of the GNSS application, higher requirements of the navigation receivers' performance have been put forward in terms of various kinds of application circumstances. It is necessary to improving the receivers' acquisition and tracking sensitivity to make it keep working in ultra weak signal scenarios such as indoor environment, tunnel et al. However, the optimization of the traditional classic acquisition and tracking structure has run into the bottleneck. The exploration of new baseband signal processing structure is needed. The massive parallel correlator technology, beside the coherent integration extension and the assisted-GPS(AGPS), is one of the international-accepted high-sensitivity technologies.This article developed an open-loop GPS signal acquisition, tracking and bit synchronization structure based on the massive parallel correlator technology on MATLAB software platform. Moreover, Viterbi Algorithm(VA) and Extended Kalman Filter(EKF) are adopted for carrier tracking and data demodulation without the traditional loop.For validating the performance of the new structure, the correct code phase, carrier frequency and data bits should be known accurately. In addition, signal with particular CNR should be created. Thus a simulative signal source is designed for create the specific GPS signal with all parameters settable. For the rough acquisition module based on the massive parallel correlator technology, the basic performance characteristics of coherent integration, noncoherent integration and the Partial Matched Filter(PMF) +FFT algorithm are analyzed first. Then the development of the rough acquisition module is demonstrated to achieve the four-dimensional search in frequency, code phase, bit edge and data bits. Simulation results have been shown that when using 80 ms coherent integration without noncoherent integration, signal with carrier-to-noise ratio(CNR)=23d BHz can be acquired; when using 80 ms coherent integration and 20 times noncoherent integration, it can acquire signal with CNR=15d BHz. The rough acquisition not only can bring about the rough estimation of carrier frequency and code phase, but also can provide the rough location of the bit edge. In the following precise acquisition module, the Chirp-Z Transform is adopted for spectrum refining so that the frequency error can be further decreased. Besides, the location of the bit edge is finally determined. It is equivalent to the pull-in and bit synchronization in tradition receivers. Simulation results have been shown that using 80 ms coherent integration and 20 times noncoherent integration, the precise acquisition module can work in signal with CNR=15d BHz as well, and the frequency error is under 2Hz.When the carrier frequency error and the code phase error have been reduced, the precise tracking of carrier phase is required for the following data demodulation. Firstly, the Viterbi Algorithm(VA) based on Maximum Likelihood Estimation(MLE) is employed to reduce the carrier phase error. Then Extended Kalman Filter is initiated using this carrier phase with small error. EKF will keep tracking carrier phase. Meanwhile, the VA is used for data demodulation. Simulation results have been shown that when CNR=18d BHz, the bit error rate(BER) is 6*10-2. Moreover, when CNR=15d BHz, the EKF can still keep carrier tracking.