Design and Implementation of Software-Defined Receiver

Acquisition in software-defined GPS receivers is computationally intensive and time-consuming, especially with the increasing demand for rapid and accurate positioning services driven by globalization and digitization. Software-defined receivers face challenges in satellite signal acquisition, which becomes more demanding with longer signals. This paper aims to enhance the efficiency of GPS signal acquisition, a critical step for determining the code phase of PRN codes and the doppler shift in carrier frequency. Real-time applications encounter issues under dynamic conditions and degraded signals, as conventional computers lack the computational power for the necessary correlation and FFT operations. To address these challenges, a method using GPU acceleration for L1 signal acquisition is proposed. This method combines signal acquisition with GPU parallel computing using the SIMT model. A CPU-GPU platform via CUDA programming allows the CPU to handle data reading and intermediate processing, while the GPU performs the core acquisition algorithm in parallel.