Using wavelet packet decomposition technique on fuzzy classify model for underwater acoustic signal recognition

A real time fuzzy logic recognition system is developed in this paper. The functions of the proposed system include two subjects, the first subject is underwater acoustic signal feature extraction by using of wavelet packet decomposition, the second subject is underwater signal pattern recognition by using of fuzzy logic model. Finally, we combine the two parts and establish a practical, real time underwater acoustic recognition system. During the feature parameter extraction stage, signal characteristic analysis and feature extraction are studied. The results prove that using the wavelet packet decomposition method for feature extraction can obtain multi-resolution characteristics. By using the unsupervised learning algorithm, the input feature data are clustering, and the centers of the data set are generated which are the templates of the feature parameters. During the underwater signal pattern recognition stage, signal identification is performed by using fuzzy logic theory. Furthermore, by defining the linguistic variables of the feature and the membership function of the fuzzy rules, a fuzzy logic algorithm is developed for the purpose of underwater signal recognition. Finally a simulation is designed using a ship signature as input data; the results have demonstrated the effective performance of proposed system.