Assessing the Performances of Hybridized GNSS Receivers in Railway Environment

Being involved in a TECH4RAIL project, the SNCF (French Railway Company) wishes to define, develop, test and prepare the qualification of location systems with low HFR (Hazardous Failure Rate) adapted to the railway environment for critical applications. This project will also contribute to national and European work (SHIFT2RAIL, CER, EUG, ESA, ...) on future location specifications (particularly safety proof) in railways. For this purpose, among the available technologies, the use of global satellite navigation systems is one of the preferred solutions because of its global coverage regardless of local infrastructures. Indeed, GNSS applications in Railways are becoming more and more frequent. So far, the focus has only been on non-safety related applications, such as in passenger information systems or freight logistic. However, for safety related application like signalling, rail navigation system must ensure the principles of Reliability, Availability, Maintainability, and Safety (RAMS) underlying its function. When moving GNSS application into the domain of safety, it appears that the GNSS only solution is insufficient for the intended application. GNSS has a long history of safety critical applications in aviation. The same basis of integrity concept for ground applications has been widely studied these last few years, with always a strong limitation found in the definition of nominal and threat models, as well as the setup of the probability of occurrence of the threats [1]. It is also stated in [2] that EGNOS signal cannot bound correctly the real positioning error due to the strong contribution of local errors that are not considered in the model. Thus, the local effects must be considered directly at the receiver level. This observation is also the basis of work on the new definition of a PNT trust level in a challenged multifrequency / multi-constellation environment with application to ground users such as in [3] for autonomous GNSS receiver. The possible contribution of GNSS/Image fusion has been studied in [4]. [2] also proposes a comprehensive review of Rail Systems and train sensors (such as IMU, Map, odometer, …) that may contribute to the position accuracy and integrity. Thus, at least the following issues must be resolved to achieve the system objective: first, the GNSS accuracy and error bounding, even in multi-constellation & multifrequency mode, are far from sufficient to identify 100% of the track changes, particularly in urban environment. Second, GNSS signals are not always present, because of underground tunnels and underground stations, like Montparnasse (Paris) or Penn Station (New York City). Third, the many types of interferers encountered during a given trajectory. Finally, when designing a certifiable GNSS-based train location system, even with inertial hybridization, odometer and map matching, the only technology which gives an absolute positioning is the GNSS. Thus, before starting any development, we must solve the initialization process of the positioning system, either inside underground stations, or inside a tunnel after a system reset. One of the industrial-grade technologies which could solve this initial problem may be SubWAVE+ technology of Syntony GNSS [5], which allows the initialization of the PVT on the correct track in “deep-indoor” environments. Afterwards, the nominal mode of the hybridized GNSS receiver can run, its main aim being to assess any track change with an extremely low error rate. In order to assess the local environment contribution to positioning errors and to prepare the development and future qualification of GNSS hybridization architectures, the SNCF involved SYNTONY in an experimental campaign conducted in a railway environment between Vitry-sur-Seine and Montereau Fault-Yonne (~170 km round trip). The environment is representative of urban and suburban conditions, including railway stations, industrial metallic buildings, masking, tunnels, multipath and probably civil interferences. The data collecting system consisted in a high grade IMU and 2 SYNTONY GNSS Record and Playback systems ECHO. Digital railway maps have also been made available. The ECHO-Recorder was configured as follows: • 3 GNSS Bands: L1-E1 / L2 / L5-E5a • Sampling frequency: 50 MHz per band • Quantization: 16bits I&Q per band • Both ECHO-R were located in the car. Narrowband RF interferences were observed on L5 and a wideband interferer on L1 with a power spectrum characteristic of a PPD. The paper will first detail the experimental campaign. The availability of raw I&Q samples will enable: - on the one hand to feed COTS receivers using the SYNTONY ECHO-P RF player in order to o compare the behaviour of several receivers based strictly on the same reference datasets o analyse the error areas and reception conditions in these areas - on the second hand, I&Q samples can directly feed SDR receiver tracking loops in order to characterize the local environment events at signal level and to quickly test the performances of events detection and mitigation techniques. Finally, potential benefit of GNSS / IMU hybridization schemes and interference mitigation techniques will be presented based on SYNTONY GNSS real-time SDR receiver SoftSpot. At the time this abstract is written, the experimental campaign has already been conducted and the data collected are ready for processing. [1] GNSS Evolutionary Architecture Study, Phase II - Panel Report - February 2010 [2] J. Marais, J. Beugin, M. Berbineau, “A survey of GNSS-based Research and Developments for the European railway signalling”, IEEE Transactions on Intelligent Transportation Systems, DOI: 10.1109/TITS.2017.2658179 [3] G. Carrie, D. Kubrak, M. Monnerat, J. Lesouple, “Toward a new definition of a PNT trust level in a challenged multi frequency / multi-constellation environment”, Proc. NAVITEC 2014, Noordwijk, The Netherlands, Dec. 2014. [4] C. David, V. Gui, A. Campeanu, C. Nafornita, G. Carrie, P. Martin, M. Monnerat, “IMFUSING – GNSS Localization in Constraint Environment by Image Fusing Techniques”, Proc. 6th CEAS 2017, Bucharest, Romania, 16-20 Oct. 2017. [5] J. Korsakissok, “SYNTONY, Game changer of the location business”, Proc. ICSS 2019, Hyderabad, India, 7-9 Feb. 2019