Robust Reconstruction Method Based on QCQP Optimization for Multichannel SAR With Closely Singular Sampling

Abstract

The conventional reconstruction algorithm allows for unambiguous recovery of the Doppler spectrum of multichannel synthetic aperture radar (SAR) system in azimuth with an accurately known channel response matrix. However, the available knowledge of the actual channel response matrix is imprecise due to channel errors. In this case, the SAR image will suffer from severe error-induced azimuth ambiguities if reconstructed by the conventional reconstruction algorithm, especially for the case of closely singular sampling. To this end, a novel method based on a quadratically constrained quadratic program (QCQP) optimization is proposed to increase the robustness to the channel errors. Accordingly, the error-induced azimuth ambiguities can be greatly suppressed by the proposed method. Furthermore, performance comparisons of the conventional reconstruction algorithm and the proposed method via numerical simulation are presented.