The Starting Point Problem of Parameters Estimation for LFM Signal Based on Newton's Method
Deng Zhen-miao · Dianzi xuebao · 2009
In this paper,the starting point problem of parameters estimation for LFM signal based on Newton's method is investigated.The DPT algorithm is used to gain the estimated value of the coefficient of frequency modulation and the center frequency of the LFM signal,and the two estimation values are used as the starting point of Newton's method.The performance analysis show that when signal-to-noise ratio(SNR) is greater than the SNR threshold of the DPT algorithm,the two estimations are in the range of the convergence region needed by Newton's method.So it will guarantee the convergence of Newton's method.The simulation results show that when SNR is greater than SNR threshold,the RMS error of the parameter estimations can reach the CRLB.The computational complexity of the algorithm proposed in this paper is low,and the algorithm is convenient for realizing quasi-real-time LFM signal processing.