Underwater multi-target passive detection based on transient signals using adaptive empirical mode decomposition

Underwater acoustic passive detection is the basis of target detection and recognition in underwater wireless sensor networks. However, with the development of noise reduction technology, the difficulty of passive detection on steady noise is increasing. Transient signals exposed by underwater targets under some circumstances are hard to be eliminated. To detect multiple quiet targets, different time scales of transient signals are studied and a multilayer adaptive separation method based on empirical mode decomposition is proposed. At first, the characteristics of different kinds of transient signals are analyzed. A mixed-signal model is established for simulation. Then, the empirical mode decomposition method is used to extract signal components from signal pieces, dividing the signal into a high-frequency part and a low-frequency part. The low-frequency part is resampled, and the complementary ensemble empirical mode decomposition with adaptive noise algorithm is used to solve the mode hybrid problem. Principle components are picked out, and start and end times of signal components are detected. Finally, the direction of arrival estimation of each signal component is realized in the average sound intensity method. In the process, the compromise between computation complexity and error is proposed to achieve online work. Experiment results show that different kinds of signals can be divided and directions of multiple targets can be well estimated.