A New Deep Learning Method for Underwater Target Recognition Based on One-Dimensional Time-Domain Signals
Song Xiaoping, Cheng Jinsheng, Yuan Gao
Abstract
How to extract effective target features from complex underwater acoustic signals and better classify underwater targets and ships has always been an important problem in the field of underwater acoustic countermeasures. Due to the complexity of underwater acoustic environment and continuous development of underwater acoustic countermeasures, the limitation of expert experience system based on traditional spectrum analysis technology is becoming more and more obvious in passive sonar target recognition. In recent years, deep learning has made remarkable progress in image recognition, speech recognition and other fields. In this paper, a new method is developed in which one-dimensional time series data is used as the input, and some popular networks such as Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) are used as deep learning models to mine intrinsic features, and Wasserstein Generative Adversarial Network-Gradient Penalty (WGAN-GP) is used to enhance the training samples data. This method has shown an excellent performance in passive sonar recognition.