| It’s inevasible to analyse LFM(Linear Frequency Modulation) signal in manyapplications. In the meanwhile, because LFM signal has a lot of advantages, it’s widely usedin many domains such as radar、wireless communication and so on. Furthermore, it plays animportant role in underwater communication and underwater acoustic detection, and it can bethe important evidence to evaluate the effectiveness of the time-frequency analysis tool. So itis meaningful to analyse LFM signal and to find the effective method to estimate itsparameters. And the technique of LFM signal parameters estimation is an important basis ofmany application domains.Firstly, the article summarizes non-time frequency analysis and time frequency analysis,which are two kinds of methods of LFM signal parameters estimation. On the basis, thearticle picks up the method by delay autocorrelation and the method by multiplying by LFMsignal from non-time frequency analysis to study and simulate, picks up short time Fouriertransform and Wigner-Ville transform from frequency analysis to study and simulate, andanalyse the advantages and disadvantages of these methods. Moreover, this article focuses onexpatiating the estimation method based on Fractional Fourier Transform (FRFT), analyses itsadvantages by theory analysis and simulation experiment, and in the meanwhile, expatiates itsdisadvantages such as the conflict of high precision and low computation. Finally, the articleaims at reducing the conflict and puts forward an estimation method which combines geneticalgorithm and FRFT. The simulation shows that the proposed method has a much betterperformance than the method by fixed step FRFT. When SNR(Signal Noise Ratio) is higherthan-5dB, the proposed method has higher precision than the method by fixed step FRFTwith nearly the same computation; the proposed method can converge to the optimal valuefaster with SNR increasing, and it is stable to a range of iterations. |
PDF Full Text Request