The Method Of High Resolution Estimation Of Multi-component Of Doppler Signal

Since the 20 th century, radar has been widely used in various fields. When the radial velocity exists, the echo signal consists of the Doppler frequency. The parameter estimation is an important aspect of radar signal processing; the direction finding also attracts people’s attention nowadays. Traditionally Doppler signals are approximate as the constant frequency signal model for signal processing in each data frame; however when there is a dramatic frequency change in the data frame, the estimation results are bound to decline. In this paper, the Doppler signal is approximately as chirp signal model; Mathematical derivation proves the reasonableness of approximation. Simulation proves the errors of the chirp signal model are less than the errors of the constant frequency model. In this paper, the methods of subspace signal processing and the tensor decomposition method are applied to the field of the chirp signal parameter estimation. The traditional signal processing method is the time-frequency signal processing, the principle of the idea is the physical meaning of the model and the subspace rotation invariant properties, the simulation proves the applicable of the algorithm.Conventional DOA estimation method is based on the constant frequency narrowband signal model, in the scene of chirp signal model the conversion to the narrowband domain is needed, this paper chooses to use fractional Fourier transform(FRFT) to convert the chirp signal to Fractional domain, then combines with the multiple signal classification(MUSIC) method, a new 2-DOA method based on the L-shaped array is proposed, the simulation proves the effectiveness of the work.This paper studies the parameter estimation and DOA of the multi-component Doppler signal and the simulation verifies the applicability of the proposed method in this scenario.The innovative work of this article is reflected as followed:(1) The traditional chirp signal parameter estimation is based on thetime-frequency signal processing method, in this paper the subspace method and the tensor decomposition are introduced into the domain of chirp parameter estimation, the simulation proves the effectiveness of such methods.(2) The 2-DOA has been a complex issue, a new 2-DOA method based on the L-shaped array is proposed and has been proved to be effective in the wideband scene and the signal processing is easier than the similar methods.