Fast GPS Signal Acquisition Using FPGA

This thesis mainly deals with fast acquisition and real-time processing in GPS receiver. Then the simulation and realization in Field Programmable Gates Array(FPGA) have been introduced in detail.In GPS receiver many digital signal processing steps must be considered in order to align the incoming signal and local signal. But one of the most important questions is how to finish the acquisition of the signal in real time. So many algorithms for acquisition of GPS signals have been developed and evaluated. But until now, the developed algorithms in current GPS receivers are not fast enough to acquire GPS signals in real time. In GPS receiver, a circular correlation costs large computation, Computation of the correlation in time domain is a very time-consuming process. But performing the correlation in frequency domain will be much more efficient..Concurrently there are two ways in searching GPS signal: sequential detecting technique and parallel detecting technique. In this article, firstly, the performance of sequential detecting technique is considered, then the simulation in Matlab and the realization in FPGA are introduced in detail. After that, the parallel detecting technique is introduced. This method firstly translates the incoming signal into frequency domain by using FFT, after finishing the correlation, the result is translated back to time domain by using IFFT. This can avoid much complex computation and it is much faster than sequential detecting technique in searching GPS signalThe XILINX Virtex II Xc2v3000 device was used in this design. In order to satisfy the resource in FPGA, a averaging method was introduced. It mainly down--samples the incoming 6000 point serials into 1024 point without affecting the acquisition performance. From the result of Matlab, althougth the SN ratio decreased a little, the acquisition result was not been affected.Lastly, the realization in FPGA of this method was introduced in detail. It shows that the parallel detecting technique is actually a good method.