Fluid flow queue models for fixed-mobile network evaluation
Document typeMaster thesis
Rights accessOpen Access
A methodology for fast and accurate end-to-end KPI, like throughput and delay, estimation is proposed based on the service-centric traffic flow analysis and the fluid flow queuing model named CURSA-SQ. Mobile network features, like shared medium and mobility, are considered defining the models to be taken into account such as the propagation models and the fluid flow scheduling model. The developed methodology provides accurate computation of these KPIs, while performing orders of magnitude faster than discrete event simulators like ns-3. Finally, this methodology combined to its capacity for performance estimation in MPLS networks enables its application for near real-time converged fixed-mobile networks operation as it is proven in three use case scenarios.