Additive noise influence on the bivariate two-component signal decomposition

Decomposition of multicomponent signals overlapping in the time-frequency domain is a challenging research topic. To solve this problem, many approaches have been proposed so far, but only to be efficient for some particular signal classes. Recently, we have proposed a decomposition approach for mul...

Full description

Saved in:
Bibliographic Details
Published in2018 7th Mediterranean Conference on Embedded Computing (MECO) pp. 1 - 4
Main Authors Brajovic, Milos, Stankovic, Ljubisa, Dakovic, Milos, Mandic, Danilo
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.06.2018
Subjects
Concentration measures
Eigenvalues and eigenfunctions
Embedded computing
Matrix decomposition
multivariate signals
Noise measurement
signal decomposition
Signal resolution
Signal to noise ratio
Time-frequency analysis
time-frequency signal analysis
Online AccessGet full text
DOI10.1109/MECO.2018.8406053

Cover

More Information
Summary:Decomposition of multicomponent signals overlapping in the time-frequency domain is a challenging research topic. To solve this problem, many approaches have been proposed so far, but only to be efficient for some particular signal classes. Recently, we have proposed a decomposition approach for multivariate multicomponent signals, based on the time-frequency signal analysis and concentration measures. The proposed solution is efficient for multivariate signals partially overlapped in the time-frequency plane regardless of the non-stationarity type of particular signal components. This decomposition approach is shown to be also efficient in noisy environments. In this paper, we investigate the limits of the decomposition efficiency subject to the signal-to-noise ratio and initial phase differences between the signals from different channels. The paper is focused on the decomposition of bivariate two-component signals.
DOI:10.1109/MECO.2018.8406053