On Petri Net-based modeling paradigms for the performance analysis of wireless Internet accesses

In this paper we model a wireless Internet access system based on the GSM/GPRS (Global System for Mobile communications/General Packet Radio Service) technology, with three different Petri net-based paradigms: Generalized Stochastic Petri Nets (GSPNs), Deterministic and Stochastic Petri Nets (DSPNs), and Fluid Stochastic Petri Nets (FSPNs). We show that all three Petri net-based paradigms provide very similar performance predictions for some of the considered realistic configurations of GSM/GPRS systems, and for some of the performance metrics of interest. This proves that FSPNs can be an adequate tool for the performance analysis of wireless Internet access systems, and allow the investigation of large system configurations, thanks to the availability of very efficient solution algorithms for FSPN models. However, we also show that in other cases (which can be clearly characterized), the performance predictions generated by FSPN models may become quite optimistic, and thus not suitable for a careful design and planning of GSM/GPRS systems. In those cases, both DSPNs and GSPNs can provide a viable modeling alternative, at least in the case of medium scale systems.