A Modified MUSIC Algorithm Based on Eigen Space.

Qian Zhao,WenJuan Liang
DOI: https://doi.org/10.1007/978-3-642-23777-5_45
2011-01-01
Abstract:Among the algorithms of DOA estimation, MUSIC algorithm which has very high resolution and estimation precision. However, MUSIC, Improved MUSIC and Novel MUSIC can not estimate the signals when the signal intervals are very small in the case of the low SNR and existence of the signal correlation. To solve this problem, a Modified MUSIC is proposed in this paper which takes advantage of the signal subspace and the noise subspace, by making them orthogonal to each other through repeatedly reconstructing covariance matrix, we can obtain the two noise subspaces and signal subspaces. By averaging previous results and reconstructing covariance matrixes of signals and noise, DOA could be estimated through spectrum function .Computer simulation proves that among three methods only Modified MUSIC can accurately estimate the correlated signals with low SNR and its side beam is lower than the Novel MUSIC.
What problem does this paper attempt to address?