Extraction Of Scattering Centers Using A Greedy Algorithm For Traffic Participants

Reflected signals from a high number of scattering points of complex traffic targets can be utilized to model their back-scattering behaviour for radar simulations. For this purpose, the vast amount of scattering points should be reduced to a few significant scattering centers to minimize computational effort. In the following step, the extracted scattering centers can be used for the simulation of the millimeter-wave automotive radar channel in realistic scenarios. In many cases, a precise CAD models of relevant traffic participants are not accessible, creating a simplified scattering model based on the relevant scattering centers will be very helpful. The scope of this work is to present a technique to generate a significantly simplified Radar Cross Section (RCS) model of typical traffic objects with a limited number of virtual scattering centers, each with own specific scattering characteristic, and how to group these scattering centers in a cluster database. The work is based on ray-tracing simulations of complex traffic object models. The scattering centers may not physically exit, but represent virtual scattering centers with a defined scattering behaviour. In this paper, a technique for extracting such virtual scattering centers from a complex 3D-model to create a simplified model is presented.