Abstract:

To address the difficulty of solving the problem of distributed sensor detection and fusion based on noise statistical information, a distributed detection fusion method for detecting underwater target radiation noise in non-Gaussian noise is proposed. Firstly, the non-Gaussian noise in the underwater acoustic environment is modeled as an Alpha stable distribution model, and the target radiated noise is modeled as a Gaussian signal. Then, a Gaussian function detector is used to convert the polluted target radiation noise by non-Gaussian noise into a signal detection model in Gaussian noise, and the parameter relationship of the noise and signal distribution between the detector output and input is formulated. Finally, using the output sampling sample of the Gaussian function detector as the detection statistic, the detection threshold and fusion rules of the distributed sensor detection fusion system are designed under the Neyman-Perason (N-P) criterion, which results in the optimal fusion of the detection results of each sensor. The computer simulation results verify the correctness and effectiveness of the method.

Key words: underwater acoustic sensor, non-Gaussian noise, Gauss signal, distributed detection fusion, Gaussian function detector