Fast Solve of Broadband Electromagnetic Scattering Problems Based on Krylov Subspace Basis Functions Combining With Compressive Sensing

While using the Krylov subspace basis functions (KSBFs) method to analyze the broadband electromagnetic scattering problem, a large amount of time is required to recalculate the KSBFs at each frequency point. To address this issue, this paper proposes an ultra-wideband KSBFs (UKSBFs) method integrated with compressive sensing. The UKSBFs, derived through Arnoldi iterative computation and partial orthogonalization at the highest frequency point within the range of interest, serve as the sparse basis. A few rows from the impedance matrix and excitation vector are extracted to form the measurement matrix and measurements, constructing a compressive sensing framework for the efficient solution of the induced current. Benefiting from only the UKSBFs of the highest frequency points need to be calculated and efficient recovery algorithm, the method significantly reduces the radar cross section time for computing broadband targets. Finally, the numerical results demonstrate the computational efficiency and accuracy of the method.