Estimation of the Base Station Position Error in a RTK Receiver Using State Augmentation in a Kalman Filter

Low-cost single frequency Real-Time Kinematics (RTK) modules have recently been released by several manufacturers. This type of receivers allows to obtain much better accuracy, reaching decimeter-level accuracy, than traditional low-cost receivers, thus opening the world of precise GNSS positioning to a new sector. However, while this type of system will provide very good relative positioning accuracy, the absolute positioning accuracy might be degraded if the position of the RTK base station is not estimated with sufficient accuracy. Any bias on the RTK base station position will introduce the same bias on the RTK rover position. This paper proposes a modification to the position estimation algorithm that includes the real-time estimation of the RTK base station position error, by combining both the Single Point Positioning Solution and the RTK solution. The algorithm is illustrated using 2 types of real data: first, for a fixed reference station using GNSS observations only, then for a moving vehicle using a sensor fusion algorithm between GNSS, inertial and odometer observations. Performance analysis shows that the bias affecting the absolute position of the RTK rover can be estimated using the proposed algorithm, decreasing the horizontal bias from a few meters to a few decimeters.