Mixed Sources Localization Based on Sparse Signal Reconstruction

Bo Wang,Juanjuan Liu,Xiaoying Sun
DOI: https://doi.org/10.1109/lsp.2012.2204248
2012-01-01
IEEE Signal Processing Letters
Abstract:In this letter, a novel mixed sources localization method based on sparse signal reconstruction is presented, which can efficiently estimate direction-of-arrival (DOA) and range parameters of near-field and far-field sources. By constructing the cumulant domain data of array which is only related to DOA parameters of mixed sources, we obtain DOA estimation of all sources using the weighted l 1 -norm minimization. And then, a mixed overcomplete matrix on the basis of DOA estimation is introduced in the sparse signal representation framework to estimate range parameters and distinguish far-field sources from mixed sources. Compared with the two-stage MUSIC algorithm, the proposed method can provide improved accuracy and resolve closely spaced sources. The simulation results show the effectiveness of our method.
What problem does this paper attempt to address?