Nonlinear wavelet denoising for DOA estimation by MUSIC

MUSIC is a widely used technique of plane wave direction of arrival (DOA) estimation, which is a problem of great interest in several applications. The performance of MUSIC degrades under low signal-to-noise ratio (SNR) conditions due to errors in estimating the data covariance matrix from finite data. It is known that linear wavelet transform can be used for denoising signals in Gaussian noise, but this method is not suitable if the noise is strongly non-Gaussian. In this paper we discuss the possibility of employing the nonlinear wavelet denoising [David. L et al., (1997)] to improve the performance of MUSIC under low SNR with strongly non-Gaussian noise. We propose the application of nonlinear wavelet denoising to the noisy signal at each sensor to boost the SNR before performing DOA estimation by MUSIC. A comparative study of the performance of MUSIC is presented for the undenoised and denoised data. Computational results are presented to show that denoising leads to significant reduction in the bias and mean square errors (MSE) of the DOA estimates and enhancement of resolution.