Crack STIT tessellations for city modeling and impact of terrain topology on wireless propagation

This work proposes to model city maps with a particular family of random tessellations: Crack STIT tessellations. This family of tessellations allow the generation of realistic city maps with a reduced number of parameters: the Anisotropy Ratio (ξ) and the Mean Building Block Area (MBBA). These city models are then used as input to a 3D ray tracing simulator to compute received power distributions. The objective of this investigation is to identify the impact of random tessellation parameters on wireless propagation. For this purpose, path loss exponent γ is estimated from received power distributions obtained by simulation for different values of ξ and MBBA pairs. Simulations results clearly show a linear dependency between path loss γ, ξ and MBBA. Path loss γ increases with ξ while it decreases with MBBA. Furthermore the evolution of the surface occupation ratio between outdoor and overall surface is also investigated. This ratio is named Free Area Ratio (FAR) in the sequel. It decreases with MBBA according to a power law. The results obtained are particularly promising as they provide a parametric relation between path loss exponent γ and terrain topology. This considerably simplifies the radio planning and dimensionning process of cellular networks.