Abstract: In order to detect the Brillouin optical time domain reflectometry signal (BOTDR) of photoelectric composite submarine cable from the noise background effectively, wavelet threshold method was proposed to denoise the real-time monitoring signal according to the characteristics of signal. Optimal parameters of wavelet threshold denoising suitable for Brillouin optical time domain reflectometer signal of the cable were determined through theoretical analysis and experimental comparison. The effect of wavelet threshold denoising was compared with the effects of median filtering method and mean filtering method. The results show that compared with both the traditional filtering methods, wavelet threshold method with optimal parameters can effectively denoise the signal. It not only can improve signal noise ratio to 14.1dB, but also can detect strain change of 100με. The study has important reference for efficient processing of BOTDR of submarine cables.