Land Clutter Data Generation Using Generative Adversarial Network

Abstract

Land clutter data is a key part in radar signal processing algorithm design. Measured land clutter data is important to radar users, but usually not enough for radar signal simulation. In this work, we measure the scattering of a typical ground object sample. Then we use generative adversarial network to learn measured land clutter data, and generate more data which can be applied to further radar signal simulation. Numerical simulations verify the validity of such method.