Design and Implementation of a Single-Frequency L1 Multiconstellation GPS/EGNOS/GLONASS SDR Receiver with NIORAIM FDE Integrity

During the last ten years the GNSS scenario has undergone very significant changes since more and more countries have started setting up their navigation systems. GPS and GLONASS are also going through their own phases of modernization with the GPS III programme and the GLONASS M and K programmes introducing CDMA signals in addition to the already existing FDMA ones. Over the next years, Galileo and Beidou/Compass shall become fully operational systems, broadcasting new signals with modern modulations and innovative services. In this ever-changing context, the availability of reliable and flexible receivers is becoming a priority and the use of software receivers instead of commercial hardware receivers is generating widespread interest in the GNSS receiver industry. The use of a multi-constellation (by now GPS and GLONASS) for PVT computation exhibits several advantages when compared to a GPS only solution. The increased number of satellites improves the constellation geometry, enhances position accuracy and increases the solution availability. The latter is important in situations when part of the sky is blocked by obstructions, so that the signals from several satellites may not be received. In such situations, adding GLONASS constellation to GPS significantly increases the availability, the accuracy and important advantages regarding the Integrity Monitoring can be reached, too. The scope of the whole work presented in this paper has been the development of a Multi constellation SDR Receiver which had to be easily upgradeable for future GNSS signals exploiting a configurable architecture to use both GPS and GLONASS signals to compute PVT by using EGNOS augmentation system, too. In the next sections the different techniques used for acquisition, tracking and navigation data demodulation of GPS and GLONASS signals due to their different multiplexing methods will be analyzed. The strategy used to handle different time and spatial reference frames, mandatory to have a real full interoperability, and the way of combining different measurements computed by different constellations are presented. As known, the common way of computing a combined GPS and GLONASS PVT solution is to solve for two separate solutions and then combining the results (with all the disadvantages in term of availability accuracy and integrity, e.g. the need of at least 4 satellites for each constellation) or, another approach, can be considering a single solution solving the system with five unknowns taking into account the time offset between the GPS and GLONASS time scales. Here the two navigation systems are referred to a common time (GPStime) and spatial frame (WGS84) computing PVT solution through a weighted least squares technique with only four unknowns. This is possible using the a-priori information concerning the offset between GPS and GLONASS systems time scales broadcast by GLONASS-M satellites. Integrity monitoring shall take into account of using different constellations simultaneously, too. In this paper a multiconstellation NIORAIM FDE algorithm is proposed and analysed. The experimental results show that the proposed algorithm is able to identify and exclude up to two satellites of the multiconstellation, if affected by a bias, before the position error exceeds the required protection levels. The last important aspect which is analysed in the paper is the error model of the pseudorange measurements for GLONASS satellites. Combining the measurements coming from different constellations is an efficient strategy only if the weights of the pseudoranges for each satellite are well defined and this is not always possible. In the case of GLONASS measurements it is often difficult to find exhaustive and precise error models; for this reason an ad hoc pseudoranges error model was developed starting from an extensive campaign of real data analysis. The model is then used for accuracy performance improvement (WLSE method) but also to properly weights each of the satellites in the NIORAIM FDE algorithm. All the proposed algorithms are implemented in a SW prototype using the SDR techniques. Signals demodulation, navigation data extraction, observables estimation, GPS corrections and/or SBAS corrections, and PVT are performed in post processing.