Resumen
The suitability of EGNSS (including EGNOS and Galileo early services) for safety railway applications has been analysed by several European projects such as Grail, Grail-2, Satloc, 3InSat, and it is currently being analysed by ERSAT EAV. ERSAT EAV in particular is addressing the following three challenges beyond the state-of-the-art: a) reuse of ETCS odometry by adding the virtual balise concept to eliminate the fixed balise along the line; b) adoption of the public EGNOS Augmentation network ?upgraded? with Local augmentation networks to fulfil the railways? requirements; c) verification and validation of alternative GNSS solutions to guarantee localization functions in areas where the GNSS signal is not available and/or subject to interference. This paper focuses on the third of these challenges, addressing the design and the implementation of a complementary positioning system (CPS) for GNSS-denied areas. This system is based mainly on the Public Land Mobile Network (PLMN) information but also incorporates additional source information (such as predefined GNSS blocked zone information and GNSS positioning information for calibration purposes, when available) in order to enhance positioning performance. This paper deals with the different steps taken to create the CPS: 1.Analysis of the alternatives for complementary positioning systems when GNSS is not available 2.Simulation and development of algorithms for the CPS in the lab 3.Creation of a prototype for the CPS based on COTS components. More specifically, after analysing the state-of-the-art technologies, the algorithms for the selected solution will be implemented in the laboratory. In order to analyse the theoretical performance of the algorithms, the environment and observables for the algorithms will also be modelled in the simulation platform. The performance of the CPS will be analysed in terms of accuracy, reliability, continuity and integrity in a way similar to the GNSS systems. Finally, the algorithms will be implemented in a prototype based on COTS components so that the CPS will not be a theoretical solution but also a real solution that could be useful for the on-board positioning function in trains when GNSS solutions are not available.