Separation of Phase-Corrupted Multicomponent Nonlinear Chirp Signal

Abstract

Multi-component nonlinear chirp signals (NCSs) widely exist in microwave remote sensing. In some applications, it is necessary to separate NCSs containing close components in the time-frequency (TF) domain. However, the phase-corrupted data may cause a defocused TF signature and prevent individual component extraction. To solve the problem, the optimization model is developed to reconstruct and decompose multi-component NCSs with phase errors and solved by an alternating iterative algorithm. In each iteration, the individual components, phase errors, and the regularization parameter are updated by the alternating direction method of multipliers, least-square-error criterion, and the matching pursuit principle, respectively. Finally, the effectiveness of the proposed method is verified by simulation and real data examples.