Research And Implementation Of Key Algorithms For PD Radar Signal Processing

Pulse Doppler Radar is a Full-coherence Radar based on Doppler effect,which can realize doppler filtering of single spectral line of radar echo pulse string spectrum and target detection.It has superior detection capability in strong clutter environment,which is why it has been widely used in national defense,aviation,navigation,meteorology,remote sensing and many other fields.Firstly,this thesis introduces the working principle of pulse Doppler Radar and the key algorithm of signal processing,then a specific digital sampling scheme is proposed based on the principle of the digital orthogonal sampling system.Combined with the "X-Band Ship Navigation Radar signal processing board" hardware platform,the echo sampling,digital down conversion,EMIF interface communication and other modules are designed on FPGA.Then,the Pulse Doppler Radar orthogonal sampling system and key algorithm modules are simulated by software such as MATLAB and Modelsim,which proves the feasibility of the orthogonal sampling module.Secondly,the hardware platform based on the Multi-core DSP is utilized to program the key algorithms of signal processing,including Pulse Compression,Data Rearrangement,MTD,CFAR,etc.After analyzing the time-consuming reasons of these algorithms,an optimization and improvement method for complex processing such as pulse compression and data transmission is proposed,and the verification of key algorithms is completed.Finally,the theory of distance correction,advantages and disadvantages of Keystone transform and Radon-Fourier transform are analyzed in detail for the range migration of pulse compression envelope of high-speed target in traditional signal processing flow.Then this thesis proposes a criterion of fuzzy number optimally searching in Keystone transform and a fast implementation method of Radon-Fourier transform based on blind speed sidelobe,which achieves improvement in algorithm performance,computational complexity,etc.And the feasibility of the proposed algorithm is verified by simulation.