An Image-Domain Signal Model for Azimuth Multichannel Reconstruction and Its Applications

High-resolution wide-swath (HRWS) synthetic aperture radar (SAR) may benefit from reconstructing signals in the image domain, for taking local characteristics of the scene into account. An image-domain signal model for HRWS SAR to reconstruct nonambiguous images is presented in this article. This model makes the idea of controlling the regional reconstruction performance and taking the advantage of nonuniform scatter distribution possible. Resolution is also preserved perfectly by the proposed model despite whatever tradeoff is made between the azimuth-ambiguity-to-signal ratio (AASR) and the signal-to-noise ratio (SNR). The model is derived on backprojection (BP) images and can deal with different squint angles and beam steering strategies, thanks to the BP algorithm. Although the reconstruction is done pixel by pixel, the computational burden, which depends on beam steering and the specific constraints in reconstruction, may not severely increase. Two methods for Spotlight SAR and two methods for Stripmap SAR are also presented based on the proposed model to verify the model and demonstrate its potential. Simulations on both the point target and extended target with different squint angles are carried out to verify the proposed methods, which also verify the proposed model indirectly.