Every protocol design, with or without quality of service (QoS), needs to be backed up by a performance evaluation. The use of real networks for this, particularly in the case of ad hoc networks, is difficult and expensive. Moreover, evaluations of this kind do not give useful results. A real network does not offer adjustable and easily controllable environmental parameters and the extraction of data is problematic.
For this reason, most performance evaluations and research in this domain are based on simulations. Simulations not only make it possible to test new technologies and new protocols while varying different parameters but also allow us to anticipate future results and potential problems with the protocol.
In this chapter, we provide a detailed description of the implementation of the OLSRQSUP protocol using the ns simulator. Next, we define the context of the simulation. Finally, we discuss the results of this simulation and offer an interpretation of the phenomena observed.
5.2.1. Use of the simulator
Before explaining the various stages of programming involved in this study, we must clarify certain points, including the background knowledge required for the different steps. The ns simulator remains the most widespread program of its kind for research purposes.
Almost all studies already published in the domain of ad hoc networks were carried out using the ns simulator, despite a recent ...