Low-cost single-frequency GNSS receivers with patch antennas can track carrier phase measurements with millimeter- to centimeter level and can therefore provide position information comparable to geodetic grade receivers. However, in order to use carrier phase measurements for positioning, one has to first resolve the integer ambiguities, which is challenging in the case of low-cost receivers because code multipath errors can be tens of meters. A new Kalman filter is proposed in this paper for determining jointly the cooperative RTK float solution of multiple rover receivers with respect to a fix base station. The measurement model as well as the state space model used in the Kalman filter are carefully designed for simultaneous tracking of multiple rowing boats. The measurement model exploits the correlation between measurements and the state parameters are chosen without redundancy (i.e. common states are only estimated once). The characteristic periodic movement of a racing rowing boat is also exploited in this paper to correct for carrier phase measurement outliers, i.e. cycle slip. Extensive tests were conducted with low-cost single-frequency GPS receivers. Static and kinematic test results show that with the new cooperative RTK Kalman filter, float ambiguities converge much faster and integer ambiguities can be correctly fixed in less than a minute. With the movement model, cycle slips are reliably corrected during the precise tracking of the boats after ambiguity resolution.     «

Low-cost single-frequency GNSS receivers with patch antennas can track carrier phase measurements with millimeter- to centimeter level and can therefore provide position information comparable to geodetic grade receivers. However, in order to use carrier phase measurements for positioning, one has to first resolve the integer ambiguities, which is challenging in the case of low-cost receivers because code multipath errors can be tens of meters. A new Kalman filter is proposed in this paper for d...     »