GPS Multipath Mitigation in Measurement Domain and Its Applications for High Accuracy Navigation

GPS receiver tracks both the satellite code and the carrier signals and decodes them into range and delta range from the satellite to the receiver. There are various sources of errors corrupting the code and carrier measurements. They are mainly: ionosphere delay, clock error, ephemeris error, troposphere delay, receiver noise, and multipath. Most of these errors can be removed by differential techniques due to their space correlation. The ionosphere delay error can also be estimated from the measurements on different frequencies. However, the receiver noise and multipath error are independent for each receiver. Since the receiver noise is white-like and can be smoothed via low pass filter, the multipath error is the main source of the error for high accuracy navigation. This is especially true for the code measurement, where the error could have a magnitude of tens of meters. This paper presents a technique for GPS multipath mitigation in the measurement domain and its application for high accuracy navigation, such as the WADGPS Starfire developed by NavCom. The technique is based on the time-correlation properties of the GPS multipath error. Two alternatives, a four-state Kalman filter and a simple two-state Kalman filter, are used for each GPS satellite. These are: GPS range, delta range, multipath error, and ambiguity. GPS code measurement and carrier phase measurement are used to drive the filter. Special techniques are used for the state initialization and relationships among different satellites. Experimental results are presented showing that the technique can not only improve the performances, such as the accuracy and pull-in time of phase based WADGPS, for stationary applications but also enhance the performances for dynamic applications.