| The Global Navigation Satellite System(GNSS) is rapidly developing. The high-precision, anti-multipath and anti-jamming navigation algorithms have become new research hotspots. Most of existing real-time software receivers are designed for ordinary civilian demand. In those receivers, the RF front-end bandwidth is narrow, and the baseband processing speed in host computers is low, which means they are unable to meet the requirements of high-precision, multipath mitigation and anti-jamming receiver algorithms.This thesis features some key technologies of wide-band real-time software receivers in order to analyze interference, multipath phenomena and other complex situations in real time and study typical scenes of navigation signal.This thesis sets up a mathematical model of RF front-end to analyze the impact of front-end filtering, sampling, automatic gain control and quantization on signal power. An RF front-end design approach is designed to reduce signal distortion. The result shows the RF front-end with wide band, high sampling rate and high quantization bits improves signal to noise ratio about 1~2 d B compared to narrow-band front-end. In addition, the wide-band RF front-end design is implemented by improving data transmission speed from RF front-end to host computer based on high-speed PCI Express interface.This thesis introduces a classic acquisition method, parallel code phase search algorithm, and analyzes the feasibility of participation of Graphics Processing Unit(GPU) in receiver baseband processing. This thesis transforms the original serial algorithm and parallelizes the operations of FFT, vector arithmetic and getting relevant peak maximum. A method of rapid acquisition of navigation signals is implemented by being adapted with the GPU software model. The result shows that the parallel acquisition architecture enhances the original CPU one 22 times on the operating efficiency and meets the design requirement of fast acquisition. |
PDF Full Text Request