Conventional radiopropagation models and no-wave equation approaches: an application of the percolation theory

Predicting electromagnetic (EM) fields radiated in quite complex scenarios is an urgent demand in many classes of problems. Several modelling strategies are available such as the Okumura-based model (OK), the semideterministic Walfisch-Ikegami model (WI) and the fully deterministic approaches based on the Geometrical Optics (GO) or on the Uniform Theory of Diffraction such as that on the basis of the "Single Knife Edges" (SKE) model. Some models are especially devoted to manage with specific geographical or urban characteristics, some others are too CPU-time demanding, others deserve a strong tuning effort before ensuring reliable predictions. In such a context Percolation Theory (PT) may be quite a promising new technique; a new model can be set-up without involving the classical EM wave equation based on the assumption that EM urban propagation can be assimilated to a fluid flowing through a percolative lattice. In this paper, different numerical solutions, ranging from the more standard radio propagation model up to the novel PT approach, are implemented and tested in a real case: GSM sub-networks in Umbria region. A discussion on the different model performances is proposed as well.