Evaluating aggregation protocols for vehicular networks in simulation

Abstract

Routing for vehicular networks has so far presented special challenges such as high node mobility and velocity, and until now there have been few experiments performed using real data. This thesis uses a dataset of 538 taxicabs driving around the San Francisco Bay Area to gather measurements about connectivity to lay down a framework with which to reason about these networks. Additionally. this thesis describes a simulation framework written in Python to perform rapid evaluation of aggregation protocols as well as simulations where a fraction of the cars are assumed to be internet access points.