摘要

Traditional GNSS software receivers usually process signals only on CPU, which cannot completely realize real-time processing, because it's difficult to execute on CPU quickly and efficiently for massive computations. Acquisition and tracking, two main parts of the receivers, contain massive correlations, FFTs and data accesses. Graphic Processing Units (GPU) provides parallel environment that matches the processing optimization requirements. This paper provides a strategy for optimizing acquisition, tracking and data accessing by using GPU. In our implementation, GPS Dual channel signals are used as signal source and the results reflect significant promotion