Skip to Main Content
This paper is mainly concerned with end to-end performance from the viewpoint of measurements using active probes. Since probes are merely trains of packets, packet-level modelling and simulation may intuitively appear to be the best solution. However, given the poor scalability of packet-level simulation, modelling of end-to-end paths at the scale defined by NGN is unfeasible. This spawns the problem of optimisation of modelling process as function of simulation duration, validity in view of packet-level interactions required by active probing, and general complexity of the model. Modeling network performance using aggregated traffic sources as well as its further analysis are the main areas covered by this paper.