Pulsar signal de-noising method based on multivariate empirical mode decomposition

In this paper, the de-noising method based on multivariate empirical mode decomposition (MEMD) is creatively proposed to put use to pulsar signal, filling the void that the previous methods based on the wavelet analysis have the limitations of choosing the basic functions, and the EMD algorithm´s bounded that it can´t process multiple signals jointly to avoid the mode mixing. MEMD is an extension of EMD, it has the ability to align `common scales´ present within multivariate data. Each `common scale´ is manifested in the common oscillatory modes in all the variates within an n-variate intrinsic mode function (IMF). These characteristics are especially suitable to process the pulsar signals of multiple channels with inhibition of mode mixing. Comparisons of the SNRs of the de-noised signals with that one generated by standard EMD method support this statement.