| Summary: | This article presents an efficient ray-tracing site-specific channel model compatible with wireless sensor network (WSN) simulator applications for smart cities. It relies on two 2D approaches dedicated to the main propagation modes; i.e. horizontal and vertical. The first is based on a pre-processing of the propagation environment, involving the calculation of an exact visibility graph. Computed from the discrete geometry concept called the super-cover model, it is improved to reduce its size and is used to very efficiently compute the most significant propagation paths in 3D. The specific WSN context, characterized by the sensor's radio range and mobility, is exploited to limit the size of the propagation environment and to pre-compute and store a set of visibility graphs, which are finally loaded and used on demand. The second approach is an over-rooftop model that re-uses the super-cover model to very efficiently extract the vertical profile containing the sensors, and proposes an original solution for the electrical field prediction where there are multiple diffractions in the transition zone. The results are validated against measurements, and show markedly better performance compared to others recent ray-tracing models. Finally, integration of the proposed overall solution for channel modeling in a WSN simulator is proposed, and interest in using such a model compared to conventional statistical models is demonstrated.
|
|---|
