Optical fiber laser hydrophone which has more research potentiality and value owing to higher acoustic pressure sensitivity, smaller size and lower difficulty of multiplexing. However the detecting capacity to low frequency signal of optical fiber laser hydrophone will be limited because of the low frequency noise such as 1/f noise and thermal noise of optical fiber laser and pumped laser. In order to suppress these noises, the iterative discrete wavelet transformation algorithm was designed which used multi-scale trait of wavelet transform. The different spectral components of underwater acoustic signal were separated and the noises below 1kHz were eliminated on the basis of the target signal amplitude would not be weakened. The measured data acquisitive from anechoic tank showed that the algorithm reduced the noise below 1kHz nearly 50dB and the Signal to Noise Ratio(SNR) is improved from 55.23dB to 84.05dB.