Quality control of the NRTK positioning with mass-market receivers

The employment of mass market receivers in a differential mode is not a standard procedure, especially into a network for NRTK positioning. This is because only few mass market receivers are able to yield raw data as output, almost no one accepts differential corrections, and for many applications is sufficient the WAAS augmentation. In actual fact, the improvements that can lead to the establishment of a network of CORSs (Continuous Operating Reference Stations) can be much higher with regard to these receivers, on condition that the raw data (code or carrier phase measurements) are used with special precautions. These receivers are very often used for the purpose of infomobility, but can also be used for precise farming. In this chapter, attention is focused on the quality control of GNSS positioning in real time. The goal is to show how networks using NRTK (Network Real Time Kinematic) positioning for different inter-station distances may also be useful in a real-time approach. The objective is to show how CORSs networks are useful for mass-market receivers, considering the accuracy required for the purposes described above. The accuracy of real-time positioning depends mainly on the type of receiver (whether it is single frequency or low-cost) and antenna (whether it is patch, mass-market or geodetic) used, as well as also the size of the network dimension (Dabove et al., 2011). Numerous experiments have been carried out on different types of networks and differential corrections, so we would like to show the results of such tests for the quality control of real-time positioning. Excellent results are also been obtained with regard to two mass-market receivers and two antennas settled on the roof of a vehicle. The purpose of these experiments was not to determine the direction of the vehicle, but to constrain the ‘network’ and, simultaneously, to filter the outliers. Particular attention has been devoted to the increasing the quality of the positioning of a C/A-code receiver in real time, and to analyzing and investigating the innovative methods tested by the authors that involve the indicators provided by the rover itself, such as the signal to noise ratio (S/N) and the redundancy of the observations. The experiments were conducted by splitting a single frequency antenna (Garmin) to both the receiver uBlox 6T and the geodetic receiver (Leica 1200), using the VRS correction and the differential correction of the Nearest station (about 20 km far away). With regard to the VRS correction, the correct fixing of the ambiguity phase is more than 60% of the trajectory and there were no false fixes. In such cases, the maximum planimetric error is less than 5 cm. Considering the positioning obtained by using float ambiguity (about 40%) the errors can be tightly controlled within the previously established parameters (S/N and HDOP). In this way, at least 50% of the trajectory has a maximum error of less than 20 cm. Using the Nearest correction, we have obtained a smaller quantity of integer ambiguity fix and, in this case, it is therefore possible to control the quality of the positioning using the HDOP and S/N parameters.