As the pulsar profile obtained through the folding of X-ray signal in a short time has low signal-to-noise ratio, a de-noising method based on lifting scheme is proposed for the X-ray pulsar signal. The model of noise reduction based lifting scheme is analyzed. And the pulsar profile is preprocessed by using the pre-denoising model. The noise of the pulsar profile is reduced based on the lifting scheme with wavelet and the adaptive thresholding function. Some groups of measure data explored by Rossi X-ray timing explorer (RXTE) are applied for simulation analysis. The simulation results show that the signal-to-noise ratio of pulse profile can be improved, and that the proposed method has fast computational speed.