Modified MUSIC Algorithm for DOA Estimation With Nyström Approximation

Multiple signal classification (MUSIC) can be seen as one of the most popular solutions to the direction-of-arrival estimation problem in array signal processing. Traditionally, this method utilizes the whole array observations to construct the noise subspace. In this letter, based on the randomly chosen sensors outputs, we exploit the Nyström method to approximate the noise subspace, and develop a low-complexity modified MUSIC version for far-field sources localization. Computer simulations confirm that the novel algorithm using only one-fourth sensors outputs provides almost the same estimation performance as its traditional version.