DTN routing in urban public transport systems

Communication is crucial to the coordination and efficient operation of public transport systems. However, deployment of infrastructure based communication systems is very expensive. Delay tolerant vehicular networks are a promising alternative since only very few infrastructure elements are required. This paper presents a DTN routing algorithm for urban public transport systems. Beginning with an analysis of node mobility, system characteristics are derived and exploited to improve routing performance. To increase realism in the performance evaluation and comparison a new approach is taken for the generation of mobility traces. A map based on real cartographic data is combined with line definitions, stops and timetables of real public transport systems. A micromobility simulator then produces large scale mobility traces which are fed into a DTN simulator. We compare various DTN routing schemes with our algorithm. Moreover, the impact of disturbances in the public transport system on the routing performance is examined. The results show that our routing algorithm can outperform previously proposed algorithms even if 20% of all vehicles are behind schedule.