Improving the Accuracy of IVC Simulation Using Crowd-sourced Geodata

摘要

We discuss the use of crowd-sourced geodata in simulative evaluations of Inter-Vehicle Communication (IVC) protocol designs. Typically, network simulation tools, which have been improved over decades of network research, are used for evaluating communication systems. In the area of IVC, however, additional challenges have to be met. Most important, the mobility of vehicles in network simulation needs to be represented accurately, e.g., using road traffic microsimulation techniques. These can be integrated with network simulation tools in order to provide a holistic view on the overall system performance. Obviously, the quality of these approaches inherently depends on the quality of provided map data. The OpenStreetMap project provides a community-maintained repository under an open license model. The available crowd-sourced geodata not only consists of road topology data but also includes fine-grained details such as traffic lights, speed limits, and even information about buildings, which represent obstacles for wireless communication. Using our Veins simulation framework, we show that this data increases the accuracy of IVC simulation.