A Broadcast Vehicle to Vehicle Communication System in Railway Environments

Current statistics in railway transportation systems show that the number of accidents is still too high, despite of huge investments in infrastructure-based safety systems. We show that a infrastructure-less cross-layer train-to-train communication system exploiting all characteristics of a pervasive computing system, like direct communication in mobile adhoc networks (MANETs), exchange of location and other relevant context information provided by multiple sensors in the trains, can reveal hazardous situations. While maritime and air transport are already benefiting from collision avoidance applications based on infrastructure-less communications and similar applications will be available for road users built on top of car-to-car communications soon, an equivalent for rail transport systems is not existing yet. In order to design such a system, a six phase work approach is conducted: Preliminary analysis and selection of an adequate frequency band, characterization of the propagation channel, MAC layer design, physical layer and finally verification of the system. During this process, context information provided by navigation systems and other sensors, e.g. position, time and speed are utilized to improve the communication.