Evaluation of Seed Selection Strategies for Vehicle to Vehicle Epidemic Information Dissemination

We consider the problem of how to identify a set of seed nodes in order to disseminate content efficiently in a vehicular network. We consider several relevant dimensions including proximity, encounters, speed, etc. and identify and taxonomize a number of candidate strategies. We comparatively evaluate these strategies using a set of real vehicular traces (Taxis in Beijing). We conclude that identifying seeds based on their speed, while eliminating redundant and isolated nodes, is the most effective approach, performing significantly better than the previously random seed strategy.