Joint Application Of EMD Denoising And MUSIC Algorithm In DOA Estimation

Vector hydrophone compared to traditional acoustic hydrophone pressure,has a lot of advantages,and the vector array which composition by Vector hydrophone can enhance the performance of the array. MUSIC algorithm play a important role of the array signal processing algorithm, but the drawback is that it is heavily dependent on the SNR, the resolution performance will decreased significantly with lower SNR.This paper present a new method based on combine EMD noising with MUSIC algorithm. The basic idea is that, in the first, denoising received signal on each receiver array element; the second, ues the data which has been denoised to MUSIC DOA estimation.The first half of this thesis studied the EMD algorithm and its application in signal denoising. The paper discusses the noise and the statistical properties of the noise by EMD decomposition. Under the assumption that the first component of the modal is noise, put forward a new method based on the statistical characteristics of EMD denoising noise. First, random Sort the first IMF which be got by EMD, and reconstruct the signal through it with the other components, averaging the cumulative results after repeat the process several times, due to the random noise characteristics, we got a signal with the SNR is higher than the original signal; and then the new component is decomposed again, repeating the operation several times; finally, obtain a signal with the noise power is greatly suppressed. This method can be simplified as:random order-Reconstruction-accumulate-averaging-and then broken down-repeat the aforementioned operation process. Simulation results show that the method in low SNR can get good performance, it provides a new idea for the low SNR denoising.The last half of this thesis studied the basic theory of MUSIC algorithm and analysis of the various factors that affect the accuracy of directional of MUSIC algorithm.Against the problem that MUSIC algorithm can not be accurately performed at low SNR, combined with new EMD denoising method based on the statistical characteristics of the noise be improved in the first half, and the simulation results show that, MUSIC spatial spectrum of the primary beam width sharpen, reduce sidelobe, can improve the accuracy of the signal source and multi-directional target resolution.