This paper develops a model evaluating the impact of mobility in heterogeneous wireless networks by the way of analytical expressions of the spatio-temporal evolution of QoS indicators. Such formulas are obtained by introducing a multi-user averaged mobility pattern named density of users. Among the set of densities, we use a Gaussian form of this quantity, which results from a modeling method based on the maximum entropy principle. Numerical results show the temporal variations of the QoS indicators and highlight the combined effects of network heterogeneity due to the presence of macro and small cells, and traffic variations due to mobility.