Many results have been published in the literature based on performance measurements obtained from simulations of Vehicular Networks (VANETs). These simulations use as input traces of vehicle movements that have been generated by traffic simulators which are based on traffic theory models. To our knowledge, no one has published any work based on actual large-scale recordings of vehicle movements. We use recordings of actual vehicle movements on various roadways. In order to enable analysis on this scale, we have developed a new VANET simulator, which can handle many more vehicles than NS-2 . To enable us to use our own simulator, we present results of a cross-validation between NS-2 and our simulator, showing that both simulators produce results that are statistically the same. We use our simulator to analyze the proposed authentication mechanism, which relies on ECDSA signatures , comparing it to broadcast authentication using TESLA . We perform our evaluations using real vehicle mobility, which we believe to be the first simulations using real vehicle mobility. Our comparison shows strengths and weaknesses for each of these authentication schemes in terms of the resulting reception rates and latency of broadcast packets.