Real-time And High-resolution SAR Imaging Processing System Based On ZYNQ MPSOC

Synthetic Aperture Radar(SAR)imaging is an important application direction of radar technology.It has the characteristics of all-weather,all-day and high resolution,and plays an great application value in military and civil fields.With the improvement of SAR system performance,traditional imaging processing method based on FPGA+DSP can no longer meet the requirements of real-time,high-resolution and miniaturization.In this paper,the design of real-time and high-resolution SAR imaging processing based on ZYNQ MPSOC is the main research content.Combined with the performance advantages of MPSOC series chips,a single-chip architecture of SAR signal preprocessing and real-time imaging processing is designed,which ensures that SAR imaging processing system meets the requirements of real-time and high-resolution,low power consumption,miniaturization,and high reliability.In addition,it can realize imaging processing of 32768 × 32768 and other sample points.Furthermore,the system functional simulation of algorithms,hardware test and performance analysis are completed.The main work of this article is as follows: Firstly,this paper proposes an efficient design method for SAR imaging signal preprocessing,which implements digital down conversion,decimation filter,Doppler center estimation,azimuth pre-filtering,and range pulse pressure processing in this order.For the high sampling rate of the front-end AD,a simplified digital down-conversion structure and parallel decimation filter module are designed to realize the spectrum shift and filter extraction function of the intermediate frequency signal,which can process 16 channels of parallel input data in one clock cycle.In addition,it can meet the requirement of real-time processing.After decimation filter,the Doppler center of the echo data is estimated by using time domain correlation method,which can be used to correct the azimuthal center frequency of the signal.In addition,for the problem of resource occupation and processing speed of traditional azimuth preprocessing,the in-situ storage method is adopted to realize the azimuth pre-filtering function,which improves the real-time of signal preprocessing.In order to meet the higher requirement of distance resolution,distance pulse pressure in signal preprocessing is implemented by using the frequency domain matched filtering method after the azimuth pre-filtering.Secondly,this paper designs the implementation plan of real-time SAR imaging processing based on single-chip MPSOC,which mainly includes the design and implementation of distance Doppler imaging algorithm,imaging data transmission and matrix transposition.On the basis of the similar structure of imaging algorithms,a high-multiplexing signal processing module "FFT+ frequency-domain processing +IFFT" is designed,which can support FFT,IFFT,motion compensation,range migration correction,Doppler frequency estimation,azimuth pulse pressure and other functions.This module can realize parallel computing of 16-channel data.In addition,AXI DMA is used to achieve effective data transmission,which improves the speed of imaging processing.Furthermore,for the problem of slow matrix transposition in traditional imaging processing,this paper designs a method based on AXI MCDMA software and hardware to realize matrix transposition.It can improve the real-time performance of the imaging system.Finally,the functional simulation of the imaging processing algorithm is implemented by VIVADO and MATLAB.Meanwhile,a hardware platform is established to test the experimental results.The working performance,resource occupation,real-time performance and imaging results of the system are analyzed.It takes 39.95 seconds for the imaging processing of 32768×32768 sample points.The experimental results show that the SAR imaging processing system designed in this paper has good stability,which can realize SAR images with three resolutions of 0.3m,0.5m and 1m.This system can also improve the speed of real-time imaging processing,increase the number of imaging sample points,and reduce the power consumption and complexity.