Adaptive minimum entropy decomposition on the time-frequency plane

In many applications, such as array processing and sensor networks, it is desirable to extract the source signals that generate the observed output signals. Some common approaches include principal component analysis, which assumes uncorrelated source signals, and independent component analysis, which assumes the independence of the underlying sources. In recent years, there has been efforts to perform source separation in the time-frequency domain since most real life signals of interest are non-stationary (A. Belouchrani and M.G. Amin, 1998). In this paper, we introduce one such component extraction approach on the time-frequency plane. The proposed approach extracts components that are well-concentrated on the time-frequency plane. In order to quantify the compactness or the concentration of the extracted components, we use the entropy measure as adapted to the time-frequency distributions. It has been shown that signals which achieve minimum entropy on the time-frequency plane are Gabor logons. Based on this idea, we propose an adaptive Gabor logon extraction method from a given set of observed signals. The proposed method extracts the most significant Gabor logons as the components using an adaptive filtering approach. The method is applied on an example data set to show the effectiveness of the component extraction algorithm