Knowledge-Aided Data-Driven Radar Clutter Representation

We use a knowledge-aided, data-driven, location-aware approach based on the RFView simulation software to model and estimate the effect of ground clutter in airborne radars. Using the RFView simulator, we produce many samples of potential clutter effects and, by employing the K-means clustering algorithm on the corresponding geographical power map given by RFView, represent them by a small number of virtual scatterers. These virtual scatterers are used for clutter estimation. We show comparable accuracy for significantly less computational time complexity to those of state-of-the-art methods using RFView realistic clutter-like data.