Empirical mode decomposition (EMD) is an adaptive nonlinear, non-stationary signal analysis and processing method. Engineering applications of EMD are primarily restricted by “mode mixing,” or information coupling between IMFs(Intrinsic mode function, IMF) such that the result of EMD decomposition cannot correctly reflect the real physical process. After summarizing existing analysis and processing methods, this paper proposes an algorithm of integrating the least square mutual information (with grid search for CEEMD decomposition), and the aliasing frequency to be amended (ensuring orthogonality between various IMF components to further suppress mode mixing, reduce the number of IMF components, and improve computational efficiency). The validity of the proposed method is verified through simulation experiments. The method is then used to extract the fault characteristic frequency of a rolling bearing micro-fault signal. Experimental results demonstrate that the proposed algorithm can obtain more accurate fault frequency characteristics with a lower total lumped number, fewer IMF components, and lower computational costs relative to traditional methods.