Title: GEURIW: GNSS Environment and user Requirements Characterization on the Danube River
Author(s): A.C. Pandele, Al. Radutu, M. Porretta, N.A. Croitoru, I.B. Stefanescu, C.G. Dragasanu, M.F. Trusculescu, M. Balan
Published in: Proceedings of the 30th International Technical Meeting of The Satellite Division of the Institute of Navigation (ION GNSS+ 2017)
September 25 - 29, 2017
Oregon Convention Center
Portland, Oregon
Pages: 1822 - 1851
Cite this article: Pandele, A.C., Radutu, Al., Porretta, M., Croitoru, N.A., Stefanescu, I.B., Dragasanu, C.G., Trusculescu, M.F., Balan, M., "GEURIW: GNSS Environment and user Requirements Characterization on the Danube River," Proceedings of the 30th International Technical Meeting of The Satellite Division of the Institute of Navigation (ION GNSS+ 2017), Portland, Oregon, September 2017, pp. 1822-1851.
Full Paper: ION Members/Non-Members: 1 Download Credit
Sign In
Abstract: For maritime and inland waterways navigation, the established method of augmentation for GNSS signals is currently the service provided by local Differential GNSS (DGNSS) IALA beacons. In order to increase coverage, this service can be effectively complemented by the augmentation provided by Wide Area DGNSS (WADGNSS) systems, such as Satellite Based Augmentation Systems (SBAS). EGNOS is the European SBAS. In its current implementation (V2), it is providing a Safety of Life (SoL) service for civil aviation applications since 2011. However, a dedicated SoL service for maritime applications is also planned for the next version of EGNOS (V3), [1]. In this context, the possible adoption of EGNOS V2 in inland waterways applications has been investigated in the GEURIW (“GNSS environment and user requirements characterization on the Danube River”) project. This is developed in the framework of the European GNSS Evolution Programme (EGEP) of the European Space Agency (ESA) and is led by the Romanian Space Agency (ROSA) and the Institute of Space Science (ISS). The project organized in three major stages, i.e.: a) The analysis of both the user requirements and the regulatory framework for a number of inland waterway applications which are currently in place along the Romanian sector of the Danube river; b) The comparison between the accuracy and integrity performance obtained through EGNOS and DGPS augmentation; and c) The identification of a preliminary multipath error model which is specific for the inland waterway environment. Within the project, the first stage was developed in close collaboration with the Lower Danube River Administration. The expert opinion of other stakeholders (e.g. the Naval Romanian Authority) was also considered to understand the operational concept and to identify the relevant requirements for the different components of the on-board navigation system. The second stage was based on a data collection campaign took place in March 2016 on a 555 km stretch of the Danube River. This almost covers the whole Romanian sector of the Danube River, between Moldova Veche and Giurgiu. Therefore, it includes a number of typical scenarios for inland waterway navigation, such as “flat” navigation (i.e. no significant obstacles along both sides of the river), port approach, or other specific situations (e.g. under a major bridge or in the proximity of a water-lock system). The collected data was classified according to the different scenarios to identify the possible impact on the navigation performance which might be due to the different propagation environment. In particular, the data was collected on board a hydrographic survey ship belonging to the Lower Danube River Administration. Two Septentrio AsteRx-U receivers were connected to a common antenna in order to collect GPS measurements and EGNOS corrections. DGPS and RTK corrections were obtained in real-time through NTRIP streams via an internet connection. The navigation performance comparison between EGNOS and DGPS includes: 1. The evaluation of the Position Error (referenced to a “true” RTK solution); and 2. The associated integrity parameters. In particular, integrity parameters consist of External Reliability Levels for DGPS (obtained through a standard RAIM algorithm) and Protection Levels for EGNOS (computed according to the MOPS). A “standalone” solution with GPS L1-only was also computed for the completeness of the analysis. For each epoch, the DGPS, the EGNOS and the standalone PVT solutions were compared with the RTK solution, which is considered as a “reference” solution. The standard deviation of the positioning errors considering the RTK “truth” is found to be 0.5 m for DGPS, 1.14 m for EGNOS and 0.7 m for the standalone solution, respectively. In contrast, the 95 percentile accuracy is: 1.6 m for DGPS, 2.5 m for EGNOS and 2.9 m for the standalone solution, respectively. It is to be noted that the data collection campaign took place close to the Eastern limit of the EGNOS coverage area, with a smaller number of monitored satellites and higher uncertainties related to ionospheric corrections. Thus, it is expected that EGNOS accuracy to increase towards the centre of the coverage area. Regarding integrity, while not conceptually identical, the Protection Levels (PLs)_provided by EGNOS have been compared with the External Reliability Levels (ERLs) provided by RAIM using standalone and DGPS augmented GPS L1. The standard deviation of the computed HERL was determined to be 1.5 m for DGPS and 3.8m for standalone, while the EGNOS HPL was 4.5 m. Additional statistics on both ERLs and PLs will be provided in the full paper. The collected data was also used to identify a preliminary multipath error model for inland waterway applications (third stage of the project). For this task, the data collected along the Danube River was complemented with additional measurements collected on board a larger ship docked in a typical port environment. In particular, a code minus carrier (CMC) combination was used for multipath error determination. Multipath error samples have been then then grouped for different “bins” of the satellite elevation angle. An exponential function of the satellite elevation angle was eventually used to “fit” the Root Mean Square (RMS) of the multipath data associated with each of the different bins. Based on this methodology, a dedicated multipath error model was determined for both ships and for static and dynamic data (four different models, overall). The largest values of the multipath error were observed in static measurements, but they never exceed 0.7 m on both ships. In dynamic conditions, the largest values were below 0.5 m. In both dynamic and static conditions, the difference between the multipath error values observed in the two ships are minimal. This seems to indicate that, at least based on the collected data, the major contribution to the multipath error values is not due to vessel structural parts but is associated with the propagation environment of the inland waterway scenario. However, this is just a preliminary conclusion that needs to be confirmed by additional measurement campaigns. As a number of receivers use code-carrier smoothing (i.e. the Hatch filter) to mitigate the effects of the multipath error on the code measurements, a multipath error model was also derived for the overall set of data considering three situations, i.e.: 1. no code-carrier smoothing, 2. 100 s smoothing constant (as recommended in the MOPS for civil aviation applications) and 3. 300 s smoothing constant. Compared to the first case (no code-carrier smoothing), multipath error values are reduced (as expected) when the Hatch filter is applied. However, in all of the three cases, the observed values for the multipath error are generally larger (up to 0.25m) than the multipath error model recommended in the MOPS for civil aviation applications. This confirms the need of developing a dedicated model for inland waterway applications, where the propagation environment is not controlled in terms of multipath and interference (e.g. see [2]). For that reason, additional measurement campaigns will be needed to either confirm or modify the initial indications provided by the proposed, preliminary, model for the multipath error. Additional details and comments on these findings and on the three stages of the project will be provided in the full paper. Reference [1]. Commission Implementing Decision (EU) 2015/1183 of 17 July 2015, “setting out the necessary technical and operational specifications for implementing version 3 of the EGNOS system,” Brussels (BE), July 2015. [2]. M. Porretta, D. Jimenez Banos, M. Crisci, G. Solari, and A. Fiumara, “GNSS Evolutions for Maritime – An Incremental Approach,” Inside GNSS, May/June 2016.