APPLICATION OF TIME-FREQUENCY HILBERT-HUANG DISTRIBUTION FOR ANALYZING OSCILLATING SIGNALS

Is devoted to studying the time-frequency Hilbert-Huang spectrum for adaptive signal processing. Time-frequency distribution is constructed using the components extracted from the original signal using the empirical mode decomposition (EMD). Empirical modes are used for calculating the analytic signal and further the time-frequency spectrum. We compute full phase of the signal, which is further used for calculating the instantaneous frequency. The paper also discusses empirical mode normalization for calculating the instantaneous frequency and the necessary preprocessing. Hilbert-Huang spectrum may be used in order to extract amplitude and frequency modulations in signals, classify signals, and find areas of energy concentration. This spectrum may be applied in order to calculate instantaneous energy density and marginal spectrum, which depends only on frequency and can be considered as an analogue of Fourier spectrum for non-stationary signals.

Authors: D. M. Klionskiy, V. V. Geppener

Direction: Informatics and Computer Technologies

Keywords: Time-frequency distribution, Hilbert-Huang spectrum, empirical mode decomposition, instantaneous frequency, empirical mode, oscillating signal


View full article