Empirical mode decomposition (EMD) is a powerful tool of time-frequency analysis． EMD decompo- ses a signal into a series of sub-signals，called Intrinsic mode functions (IMFs)． Each IMF contains different frequency components which can deal with the nonlinear and non-stationary of signal． Complete ensemble em- pirical mode decomposition (CEEMD) is an improved algorithm，which can provide an accurate reconstruction of the original signal and better spectral separation of the modes． The authors studied the decomposition result of a synthetic signal obtained from EMD and CEEMD． The result shows that the CEEMD has suitability in spec- trum decomposition time-frequency analysis． Compared with traditional methods，a higher time-frequency reso- lution is obtained through verifying the method on both synthetic and real data．