Research On Time-domain Imaging Algorithm For Bistatic SAR Based On Wavenumber Spectral Splicing

Bistatic Synthetic Aperture Radar(Bi SAR)can flexibly realize forward-looking and side-looking area imaging through transceiver-receiver splitting,spatial configuration and image processing,and has a wide application prospects.The image processing algorithm plays a key role in reducing the processing time and maintaining the image quality.However,there are large differences in flight speed and direction between the receiving stations of azimuth-variant Bi SAR,more complex echo models,and more difficult to build high-precision imaging algorithms,especially for forward-looking observations.In imaging processing algorithms,time-domain algorithms are applicable to various spatial configurations and have higher practical value.However,it has a high computational complexity and is prone to time-out when the imaging scene is large.Therefore,it is very important to study fast time-domain imaging algorithms.For this reason,this thesis studies and validates the proposed fast time-domain imaging algorithm for azimuth-variant Bi SAR.The main work and innovations are as follows:(1)The complexity and accuracy of Backprojection Algorithm(BP)and the reasons for its applicability to various configurations are analyzed.The calculation acceleration strategies and error sources of Fast Backprojection Algorithm(FBP)are studied,which lays a foundation for improving FBP algorithm.(2)An offset model of the waveband spectrum of the sub-aperture image in the uniform coordinate system is derived,and the splicing of the waveband spectrum of the sub-aperture image is achieved accordingly.(3)A Fast Backprojection imaging algorithm based on wavenumber spectral splicing(WFBP)is proposed.By establishing a uniform projection coordinate system,the projection operation to the final coordinate system during sub-aperture splicing is avoided and the algorithm is accelerated.(4)A simplified method for dividing the ground polar coordinate grid is proposed,and a projection skew model from two stations to the grid under the non-stop model is derived.By reducing the number of sub-aperture imaging backprojections and improving the accuracy of the skew model,the redundant calculation is reduced and the imaging accuracy is improved.(5)The phase factor of forward-looking image echo compensation in shift mode is derived.By establishing a uniform imaging coordinate system and splicing the waveform spectrum of the sub-aperture image,azimuthal focus is completed,which improves the computational efficiency of the algorithm and reduces the error of the sub-aperture image superimposed on the final image.The model and algorithm presented in this thesis are validated in the imaging simulation of azimuth-variant Bi SAR,and their validity and efficiency are verified.