A growing number of wireless communication systems are being installed on trains to improve safety and passenger comfort. However, most communication antennas are located on the roof of the train, close to the aerial supply lines and the pantograph-catenary interface. Due to the recurrent loss of contact at this interface, a high difference of potential inevitably generates frequent electric arcs, resulting in transient radiated electromagnetic disturbances sensed by the communication antennas. In a context where the railway industry is looking for new wireless technologies to replace the existing ones, an essential criterion for choosing one technology over the other is the resistance to the transients generated by the pantograph-catenary contact. In this paper, we propose a test methodology that fills a gap in the electromagnetic compatibility literature since the immunity test standards are mainly dedicated to product validation and not to the evaluation of communication systems. Our methodology has been evaluated through experiments based on a common resilience metric that allows comparison between two different communication systems used in railways: LTE and LoRaWAN.