Parameter estimation of sinusoidal harmonic embedded in colored noise: a new cross-spectral SVD-LS approach with cleared false peaks of spectral estimation

The extended characteristic equation methods are widely used by most harmonic retrieval approaches, because they can effectively improve the spectral resolving power of harmonic retrieval, and at the same time can also suppress colored noises. While the false peaks caused by these methods disturb seriously to distinguish the false peaks from the harmonic signal peaks. This paper has successfully solved the problem by studying the singular value decomposition (SVD) natures of the cross-correlation matrix. Simulation results show that this new cross-spectral SVD-LS (NCSVD-LS) approach can clear spectral false peaks