Published online by Cambridge University Press: 18 December 2014
The methodology chosen greatly influences the quality of performance evaluation. Field operational tests (FOTs) are being conducted in many areas and provide some very helpful first results. Yet, large-scale experimentation is conceptually complicated or even infeasible – just consider the possible number of parameter configurations that need to be tested. Thus, simulation is in most cases the method of choice for performance studies. This chapter studies how such simulations should be performed in the field of inter-vehicle communication (IVC) and which tools are available to aid researchers. One of the key concerns is the correct and realistic modeling of the vehicles' mobility. Besides that, the “correct” choice of the scenario has a strong influence on the expressiveness, the validity, and the comparability of simulation experiments. This chapter also studies the impact of radio signal propagation models, the influence of the human driver's behavior, and suitable metrics for finally assessing the performance.
This chapter is organized as follows.
• Performance measurements (Section 6.1) – In this section, we discuss strategies and techniques for performance evaluation of vehicular networking applications and protocols in general. In particular, we outline recent measurement campaigns, including the tools used, and give a high-level overview on simulation techniques.
• Simulation tools (Section 6.2) – This section covers typically used tools for simulating vehicular networks. We introduce the necessary network simulation as well as road mobility simulation and summarize integrated frameworks used for more holistic approaches to simulation of vehicular networks.
• Scenarios, models, and metrics (Section 6.3) – We study the scenarios, models, and metrics needed in this section. It makes a huge difference whether the protocols have been studied in the correct environment and using models accurately representing the realistic behavior of wireless communication channels and vehicular networking protocols. We summarize this discussion with an overview on metrics that can (and should) be used for the final performance assessment.