Study On Broadband LFM Signal Parameter Real-time Estimation Method Based On TMS320C6678

In various fields of radar,underwater acoustic,communications,it's widespread to estimate the parameters of chirp signal(Linear Frequency Modulation,LFM).Basing on project background,this paper realize LFM signal parameter measuring and estimation on TMS320C6678 hardware platform,it is more effective to estimate the parameters of LFM signal than traditional platforms.This paper forward a compositive LFM signal detection and parameter estimation method which is suitable for TMS320C6678 hardware platform,which is combined of Short-time Fourier transform,difference of boxes filtering,long sequence FFT,this comprehensive method has been realized on the TMS320C6678 hardware platform,and it is proved that this method has good performance,the estimating of bandwidth,pulse length and PRF is very fast.For accurate estimation of LFM modulation slope,this paper pick up a parameter estimation method based on dichotomy which is developed from delay autocorrelation algorithm and demodulation-test algorithm.It has almost the same performance with demodulation-test algorithm,but has less calculation above the SNR of-5dB.By comparing with other method,this paper proving this methor has better real-time performance and better estimation accuracy.The fast RAM is limited,it can not run long sequence FFT on TMS320C6678 hardware platform,to solve this problem,this paper forward a long sequence FFT method based on periphery RAM,it has almost the same result with MATLAB.In order to realize this comprehensive method on the TMS320C6678 hardware platform,designed and implemented a pipelined parallel processing structure,with the the help of PC and other equipment,this method has been initially realized..