DOA Estimation Based on MN-MUSIC Algorithm
Xiaochuan Wu · Journal of Information and Computational Science · 2014
By virtue of the sparse and redundant representation theory, estimating the target azimuths is equivalent to search the non-zero supports of sparse signals. In this paper, a new Direction-of-arrival (DOA) estimation algorithm called the Mixed Norm Multiple Signal Classication (MN-MUSIC) based on sparse representation theory is established. The process of searching supports mainly consists of two steps: the mixed norm and the augmented signal subspace tting. The former step aims to nd the partial non-zero supports and the latter is used to get the rest parts of the supports. Especially, the hybrid algorithm is able to make up for deciencies between compressed sensing and spatial spectrum estimation methods. Compared to the most successful sparse recovery algorithm l1-SVD and the state-of-art CS-MUSIC algorithm proposed in recent years, we demonstrate via simulations the improved resolution performance and robustness to noise of our algorithm.