Wavelet denoising techniques for DOA estimation problem

Wavelet theory has been widely used in video compression, human face recognization, signal processing, and so on. Wavelet denoising is one of the most important techniques in wavelet theory. The application of wavelet denoising in DOA estimation problem has not been systematically studied till now. We have proposed a modified version of wavelet denoising, wavelet array denoising (WAD), for MUSIC and ESPRIT algorithms. The proposed methods denoise the snapshots of each sensor in parallel, which can been viewed as preprocessing procedures to harness the performance degradations of conventional DOA estimation methods. Computer simulations have shown that with WAD the MUSIC and ESPRIT methods can present performance improvements, especially in cases of low SNR and correlated noises.