Design Of Real-time SAR/GMTI Signal Processing For UAV-borne Radar

The UAV-borne radar system can not only carry out high-resolution imaging in the detection area,but also detect the moving target in the area,which has a great development space in the military and civil fields.Due to the small size of UAV,it is seriously affected by airflow and its motion is very unstable,which requires a high motion compensation algorithm.Therefore,SAR/GMTI real-time signal processing has the characteristics of large amount of echo data,complex algorithm calculation and high real-time requirement.How to design and implement a high-performance real-time signal processing system suitable for the unmanned aerial vehicle platform has become one of the hot research topics in signal processing.The real-time signal processing module designed in this thesis,as the core of an unmanned aerial vehicle radar system,is composed of three DSP chips and one FPGA chip.By mapping the improved algorithm to DSP,it can meet the requirements of SAR/GMTI real-time signal processing.The main contents of this thesis are as follows:Firstly,an improved R-D imaging algorithm based on range-Doppler(R-D)strip SAR imaging algorithm is proposed,and the software implementation method of the improved algorithm is given.The realization process includes motion compensation,two azimuth frequency estimation,migration correction,azimuth compression,azimuth equal-spacing interpolation and geodetic correction.Then,based on the Polar Format Algorithm(PFA),a PFA Algorithm,which uses dot linear phase instead of interpolation to achieve range resampling,is proposed,and the software implementation method of the improved Algorithm is given.The specific implementation process includes motion compensation,PGA estimation,azimuth interpolation,frequency estimation and distance geometric correction modules.At the same time puts forward a PGA estimation method based on the piece of L2 within the memory space,this method can be used in the distance to the continuous data,reduce the consumption brought about by the matrix transpose,at the iterative phase estimation will all need to compute the bearing to the data prior to L2,reads to prevent,compared with the original PGA method,computing speed increased by 37.5%.Then this thesis studied the key technology of moving target detection including clutter cancellation,constant false alarm detection and DBS imaging principle,design to achieve the suitable for jagged PRF and fixed PRF multiple scanning methods of GMTI point trace formation processing and DBS imaging,the specific implementation process including motion compensation,bearing FFT,constant false alarm detection,parameter estimation and point trace condensation module,etc.Finally,for the communication between multi-chip DSP and FPGA,a SRIO interrupt service program design method is presented,and the data transfer and inertial navigation parameters compensation are accomplished in real time during the interrupt.In addition,in order to facilitate the upgrade and maintenance of radar system equipment,this thesis also proposes a D based on network port.After many flight tests,each module of the UAV SAR/GMTI real-time signal processing system has been fully verified.The obtained SAR image and the tracking information of the moving target have been analyzed,and the resolution and other indicators meet the requirements of the project,which verifies the feasibility of the design.