Abstract: The magnitude error and phase error(MEPE) in the system transfer function of the stepped-frequency synthetic aperture radar(SAR) introduces periodic MEPE in the synthesized wideband signal,resulting in periodic grating lobes in the high-re-solution range profile(HRRP).After subsequent SAR imaging processing,these periodic grating lobes appear as false targets in the SAR image,seriously affecting SAR image target detection and recognition.Therefore,the GLS algorithm based on SAR images has been proposed,which is based on the assumption of point targets and effectively suppresses image grating lobes by utilizing strong point targets in SAR images.However,for non-point target scenes,existing GLS algorithms based on SAR images can cause image defocusing while suppressing image grating lobes.Therefore,this paper proposes a GLS algorithm based on target information differences,named the target information difference method,which is not based on the assumption of point-shaped targets.By estimating the information difference between the ideal HRRP and the actual HRRP of the target after wideband synthesis,it can robustly estimate the periodic MEPE in the synthesized wideband signal.By compensating for this periodic MEPE,the algorithm can suppress the range grating lobes to the SAR image background level.By comparing the performance of different GLS algorithms,it is found that the proposed GLS algorithmis less affected by the image signal-to-noise ratio and is suitable for both non-point target and point target scenes,thus having obvious advantages comparing to existing GLS algorithms.Experimental data processing results demonstrate the effectiveness and superiority of the proposed method compared to existing GLS algorithms.

Key words: Grating lobe suppression, Stepped frequency synthetic aperture radar, Background level, Broadband signal, Image defocus