SPRITE: A New Sparse Approach for 3D High Resolution RCS Imaging

The Radar Cross Section (RCS) characterizes the scattering power of a target exposed to an electromagnetic field. Its analysis and its control are important in many applications. 3D imaging is a suitable tool to accurately locate and characterize in 3D the main contributors to the RCS. However, this is a non-invertible Fourier synthesis problem because the number of unknowns is larger than the number of data. Conventional methods such as the Polar Format Algorithm (PFA), which consists of data reformatting including zero-padding followed by an inverse fast Fourier transform, provide results of limited quality. In this paper, we propose an original new high resolution method, named SPRITE (for SParse Radar Imaging TEchnique), which considerably enhances the resolution of the estimated RCS maps. It is based on a regularization scheme that accounts for information of sparsity and support. The solution is then defined as the minimizer of a penalized and constrained criterion. Optimization is ensured by an adaptation of the ADMM (Alternating Direction Method of Multiplier) that is able to quickly perform calculations using fast Fourier transforms. Finally, the method is evaluated on real data. Compared to the PFA, the resolution is significantly increased, leading to a better RCS analysis and control.