We are interested in analyzing the performability requirements of mobile communication systems by using model checking techniques. We model these systems using a high-level formalism derived from the π-calculus, for considering stochastic, timed, deterministic or indeterministic behaviors. However, in the π-calculus, the basic communication primitive of systems is the synchronous point-to-point communication. However, mobile systems that use wireless networks, mostly communicate by local broadcast. Therefore, we first define the broadcast communication into the π-calculus, to better model the systems we study. We propose to use probabilistic and stochastic versions of the algebra we have defined to allow performance studies. We define a temporal version to consider time in the models. But the lack of tools for analyzing properties of models specified with π-calculus is a major obstacle to our work and its objectives. The definition of translation rules into the PRISM language allows us to translate our models in low-level models which can support model checking, namely discrete time, or continuous time Markov chains, timed automata, or probabilistic timed automata. We chose the PRISM model checker because, in our best knowledge, in its latest version, it is the only tool that supports the low-level formalisms that we have previously cited, and thus, makes it possible to realize complete performability studies. This approach allows us to overcome the lack of model checkers for our models. Subsequently, we apply these theoretical concepts to analyse performability of mobile wireless sensor networks.