With the proliferation of analog and digital electronic systems in the current means of transport, the EM (Electromagnetic) environment becomes richer and richer in all kinds of signals and, therefore, it becomes more difficult to characterize. In this thesis, we focus on a particular digital system: the GSM-R (Global System for Mobile communications - Railways). It is the new digital radio communication system under deployment on the European rail network in order to ensure the interoperability of high-speed trains in Europe. Then, all countries in Europe will use the same system, which will facilitate the movement of cross-border trains from a country to another one. In the railway environment, the GSM-R is subject to different EM noise sources, including transient EM interferences coming from the sliding contact between the pantograph and the catenary. These disturbances cover wide frequency bands including, potentially, those of the GSM-R system. We propose a classification method for predicting the effect of transient EM disturbances on the quality of GSM-R transmissions. This classification method could be implemented on trains where it could identify and locate critical areas for the quality of GSM-R transmissions along the journey. From the point of view of standardization, these research works could contribute to the necessary evolution of equipment and methods defined in EMC standards in order to adapt to the new problems arising from the multiplicity of wireless communication systems and protocols employed nowadays in the world of transports.
