Abstract: In order to extract the useful weak pulse signal in drilling fluids and recognize it while logging during the implementation of MWD,an effective extraction and algorithm for pulse signal was studied.The morphological filtering algorithm has been used to extract the pulse signal from drilling fluids,using the expansion and corrosion a lgorithm to design the signal morphology filter.According to the synchronous characteristics of the downhole signal with Manchester encoding,an algorithm based on a signal equal segmentation in cycle to recognize the time when starting showing pulse signal in drilling fluids accurately was presented.A waveform recognition algorithm with a pattern similarity was used to identify the pulse signal in drilling fluids.Simulation results showed that a morphological filtering algorithm can effectively extract the pulse signal from drilling fluids,the pulse signal that has signal-to-noise ratio of 20 dB can be extracted.The application of pattern-similarity waveform recognition algorithm can recognize effectively the pulse signal in drilling fluids based on Manchester encoding,the recognition range of weak pulse signal from 0.034 5 MPa to 0.069 0 MPa.According to the statistics and analysis of bit stream data,the identification algorithm can make the error rate less than 1.5%.The results showed that a morphological filtering algorithm and a pattern-similarity waveform recognition algorithm can accurately extract and recognize the pulse signal in drilling fluids,improve decoding accuracy,and meet the requirements of engineering applications.