In recent years, GPS technology and embedded technology have a good application in navigation, positioning, measurement and other fields. In order to research and analyze the performance of three kinds of common positioning algorithms (least square method, weighted least square method and kalman filter), for different receivers and considering the performance, cost, operation efficiency, stability, and other aspects of the problems, this paper designs a GPS positioning system based highperformance ARM. The experiment uses two different kinds of receivers(NEOM8T,LEA6T). The practical operation results show that the three filtering algorithms can improve the positioning accuracy of GPS, but it has different degrees of optimization performance on the receiver with different performance. This study provides an effective reference value for selection of location algorithm in engineering implementation, at the same time, GPS positioning system with highperformance provides a good base of hardware and software for differential positioning technology, carrier phase measurement, integrated navigation and other cuttingedge technology research, and has important significance in the field of satellite navigation and positioning.