The dyadic multi-resolution analysis (MRA) has a drawback, that is, there is frequency distortion in the high frequency subband, and frequency distortion is happened more serious in the process of decomposition of the dyadic wavelet packet transform. This paper presents a contribution to the 3/2 wavelet packet transform. The new algorithm includes: the method of decimation; the structure of signal processing; the computations of the frequency band width of the subband; the selection of scaling functions, wavelet functions and the corresponding filters. The simulation of numerical value and EEG example proves the algorithm can solve the frequency distortion in the process of decimation in each level.