Study On Doppler Signal Fast Fourier Transform System Based On FPGA

Optical Doppler(OD)technology has wide range of applications.Its demodulation methods are becoming more and more mature,especially the secondary applications supported by system software becoming increased,such as industrial computer-based applications.In recent years,with the development of FPGA(Field-Programmable Gate Array)technology,Doppler signal demodulation based on FPGA has become a practical need.Due to the design flexibility and parallel running speed of FPGA technology,it is almost a necessary choice to meet the requirement of the system miniaturization and low power consumption.In this paper,the basic Fast Fourier Transform(FFT)algorithm was employed to realize Doppler signal demodulation in FPGA environment.It was a part of the development work of solid-state laser Doppler velocimeter to carry out fast Fourier transform on FPGA platform for ideal and noise-added Doppler signal.The major achievements of this thesis were:(1)According to the current hardware characteristics of FFT algorithm and FPGA,base-2 algorithm was selected as the core algorithm to design,by which the FFT rotation factor is generated and generated by Matlab language and formed a.mif file,which was directly stored in the ROM of FPGA chip and programmed as a subroutine.(2)The FFT algorithm was programmed in FPGA environment from the basic principles and the modulus including: Ethernet transmission module(according to Ethernet UDP/IP transmission protocol,complete code writing and functional simulation),SDRAM storage module(according to the working principle of SDRAM to achieve data storage),FFT algorithm module(according to the FFT algorithm timing sequence to complete the control signal of data transmission,to drive the FFT algorithm for data processing,Then output the complex data after the Fourier transform,and finally take the modulus of the complex data);LCD spectrum display module(display the spectrum of the final results)and other parts.(3)Noise was added to the Doppler signal and its accuracy was tested under different sampling points and different signal-to-noise ratios.The relevant results showed that it took41.50 us to complete the 512-point Fast Fourier Transform,82.52 us to complete the 1024 points,and 164.50 us to complete the 2048 points.By analyzing the experimental results after adding noise,if the deviation degree of signal noise is controlled within 0.08(SNR is 16 d B)during FFT,the ideal processing result will be obtained.