Assessing the impact of obstacle modeling accuracy on IEEE 802.11p based message dissemination

Deploying real IEEE 802.11p vehicular network testbeds is a challenging but difficult option for most researchers. In these cases, the research community relies on simulation tools to test their protocols. However, since simulation accuracy is a critical issue, real testbed results should be used as a reference to improve simulation behavior. Our proposal adjusts common propagation models to mimic samples taken from real environments, and it uses a building aware model to achieve as much accuracy as possible in urban scenarios. We evaluate the performance differences obtained with this model against other usual simulation schemes, like line-of-sight propagation models, models where no building blockage is taken into account, and models where propagation is only allowed along streets, achieving differences of up to 70% in some measurements. Going a step further, the model is used to study the radio propagation behavior along different city layouts, showing that the actual building layout is one of the key factors affecting protocol performance in urban environments.