Characterisation of resident space objects using multistatic interferometric inverse synthetic aperture radar imaging

This paper illustrates an interferometric inverse synthetic aperture radar based on three-dimensional imaging and rotational velocity estimation from a multistatic setup with three ground-based radars with non-orthogonal baselines. The radar backscattering along the orbit is simulated based on physical optics approximation. The paper illustrates a novel shape-dependent and rotation-invariant feature called normalised eigenvalues for classification. A database of interferometric point clouds is generated using radar back scattering at different aspect angles and classified using various classification algorithms. The classification accuracy at varying SNR levels was evaluated using Monte Carlo simulations.