High-Resolution Grid-Independent Imaging for Terahertz 2-D Synthetic Aperture Radar with Spatial Under-Sampling

For the purpose of two-dimensional (2-D) imaging in the Terahertz (THz) near field through 2-D synthetic aperture radar technology, Fourier transform (FT) is one of the most popular imaging ways. However, FT-based algorithms would encounter performance loss either when spatial sampling is under Nyquist frequency or there are off-grid scatterers in the scene of interest. Therefore, by exploiting the theory of matrix enhancement and continuous parameter estimation, we propose to use matrix enhancement and matrix pencil (MEMP) method and matched filter to deal with arbitrarily located scatterers when spatial under-sampling is adopted. Through constructing a specifically expanded matrix, the information of the scatterers involved in the small data set can be enhanced. Then, highresolution grid-independence 2-D imaging can be achieved by the combination of MP and matched filter. Simulation results verify the effectiveness of the proposed algorithm.