Study On Parameter Estimation Algorithm For Multiple Sinusoidal Signal Under Noising Background

Multiple sinusoidal signal parameter estimation has been widely used in radar detection,smart antenna, bridge vibration testing, electronic communications technology and other fields. In nature, it is inevitable that the signal contains noise in the process of acquisition and transmission, therefore, extracting the parameter of multiple sinusoidal signal has important theoretical significance and practical application value under noising background.This paper studies the theory of the DFT algorithm, the ML method, the SVD method,the MUSIC algorithm, and the Classical Prony algorithm. The estimated performance of the algorithm is compared through the simulation experiments. Based on the summary of the above algorithm, in the view of the multiple sinusoidal signal which mixed with noise can't give the effective parameters estimation precision, this paper proposes the improved algorithm based on MUSIC algorithm and classical Prony algorithm. The main research work is as follows:An improved parameter estimation algorithm based on MUSIC algorithm is proposed.This algorithm has the higher frequency estimation precision through the iteration of the frequency in the peak. At the same time, according to the theory of optimum and the least square method, estimating the amplitude and phase. The simulation results show that the proposed algorithm effectively avoids the possibility of regarding pseudo peak of the strength slightly weaker as the frequency estimation values, improves the estimation precision of the frequency and obtains the effective estimation values of the amplitude and phase.A new improved Prony algorithm of the multiple sinusoidal signal frequency estimation is presented. Based on the Classical Prony algorithm, the proposed algorithm constructs a new sequence, then it establishes a new kind of Prony polynomials, finally, it obtains the high-precision frequency estimation for multiple sinusoidal signal. The simulation results show that the performance of this algorithm for frequency estimation is stable, in case of the low signal noise ratio (SNR), the improved Prony algorithm has higher accuracy of frequency estimation.Based on the improved Prony algorithm, this section presents a new multiple sinusoidal signal amplitude and phase estimation algorithm. According to the equivalent relations of the multiple sinusoidal signal in time domain, using the frequency values of the improved Prony algorithm,then the multiple sinusoidal signal amplitude and phase estimation values is obtained. Simulation results prove that the estimation of the proposed algorithm is better in the case of high SNR.