TrueNets: a Topology Generator for Realistic Network Analysis

The availability of realistic topology generators is a key component in the study of network performance. This work describes a new approach for realistic generation of topologies, named TrueNets, that uses open data provided by public administrations and crowd-sensing efforts for populated areas, maps, altitude of land and buildings; TrueNets estimates link performance with classical propagation models and produces annotated topologies of networks that can actually exist in the selected areas, thus providing not only an abstract tool for performance evaluation, but also a design tool for planning. We use TrueNets to model distributed mesh networks and we show that the generated topologies differ substantially from state-of the-art synthetic generators.