GNSS/IMU/5G fusion positioning based on factor graph in urban environment

Abstract Multi-sensor fusion is the main means of smartphone positioning in urban environments, among which the integration of global navigation satellite system (GNSS) and inertial measurement unit (IMU) has been widely studied. The traditional Kalman filter algorithm has the problem of difficult equation reconstruction when a sensor fails, the factor graph model is used to improve this problem. In addition, GNSS signals are easily affected by environmental interference, and INS has the problem of error accumulation. Therefore, this paper introduces fifth-generation (5G) unit vector observation information into the model, proposes a GNSS/IMU/5G fusion positioning algorithm based on factor graph, and verifies the model performance through multiple groups of dynamic experiments. Experiments show that compared with the traditional Kalman filter algorithm, the fusion algorithm based on factor graph has better performance. After adding the 5G factor, the positioning accuracy of the experimental equipment is improved by more than 25%, and the positioning accuracy inside the tunnel is kept within 5 m. The proposed algorithm effectively improves the stability and continuity of the positioning results in the urban environment.