Noise cancellation on low-frequency signals using Empirical Mode Decomposition

In this study, the noise cancellation problem on noise corrupted low-frequency signals by using the Empirical Mode Decomposition (EMD) method is considered. For this aim, the Intrinsic Mode (IM) functions of the low-frequency signal corrupted by white Gaussian noise are obtained by applying EMD on this signal. Savitzky-Golay filter and Least Squares Support Vector Machine (LS-SVM) regression are separately applied to the signal reconstructed using the low-frequency ones of the IM functions, and the estimation performance of the original noiseless signal is examined. It is observed from the simulations that a satisfactory result is achieved via LS-SVM regression.