Study On Technology For Small Satellites SAR Real-time Imaging

Synthetic Aperture Radar(SAR)system has the advantages of full-time,all-weather,high resolution and long range.In recent years,it has been widely used in various military and civilian fields such as marine monitoring,resource exploration,and battlefield reconnaissance.Small satellite SAR has the advantages of small size,light weight,short development cycle and flexible launch.It can be used for network formation flight and emergency response.Space-borne SAR imaging algorithms have problems such as large computational complexity and complex algorithms.It is difficult to guarantee real-time performance in a complex environment on the star.This paper mainly studies the parallel optimization of space-borne SAR imaging algorithm and the rapid development of FPGA algorithm.Firstly,the geometric model and echo model of space-borne SAR imaging are expounded,and the two-dimensional high resolution of SAR is theoretically analyzed.At present,there is a contradiction between the high resolution and wide swath requirements of space-borne SAR.This paper uses multi-mode SAR technology to generate images with different resolutions and mapping bandwidths to solve this contradiction to some extent.Then,the characteristics of the four modes of SAR adopted in this paper are compared and summarized from three aspects: SAR imaging geometric model,Doppler history of signal and azimuth resolution.Based on this,the system of space-borne multi-mode SAR is studied.parameter.Then,for the multi-mode SAR imaging algorithm,the frequency domain imaging algorithm is studied,including Range-Doppler algorithm,Chirp-Scaling algorithm,nonlinear Chirp-Scaling algorithm and time domain imaging algorithm.And the imaging precision and parallelism of the algorithm are considered comprehensively,the BP time domain algorithm is selected as a multi-mode space-borne SAR imaging algorithm.Then,the backward projection algorithm is studied in the aspect of parallel improvement.When using the Cartesian coordinate system,there is no need to interpolate and no point-by-point projection.In the sub-image fusion stage,the Scale Invariant Feature Transform(SIFT)method is adopted to calculate the complex number.Sub-images are merged.Finally,a fast backward projection method based on SIFT sub-image fusion is proposed.The effectiveness of the proposed method is verified by measured data.Finally,based on the characteristics of BP algorithm,the grid-based task-level parallelism design and data-level parallel design are selected,and then studies the rapid development of FPGA for backward projection algorithm.After analyzing the shortcomings of traditional FPGA development,the modeling process of backward projection algorithm in Simulink is studied based on MATLAB high-level language.The module simulation results are compared with the MATLAB results to verify the correctness of the Simulink simulation results.After the fixed-point analysis of the Simulink module of the established BP algorithm,the module is converted into the Verilog code of the FPGA,and be verified on the Virtex-7 690 t chip,and the realization of the backward projection algorithm of the single-chip FPGA is completed.The rapid development of FPGA based on MATLAB is helpful to study the real-time imaging algorithm of SAR imaging system,and lay a foundation for building a special module library of SAR imaging algorithm to realize the rapid realization of FPGA for SAR imaging algorithm.