Application of Graph Signal Sampling in Underwater Distributed Cooperative Detection

In this paper, graph signal processing is introduced into the research of underwater distributed cooperative detection. Under the limited bandwidth of underwater acoustic channel, the selection of the optimal nodes that can simultaneously interact with the fusion center is viewed from the perspective of graph signal sampling, and the graph signal sampling method based on error minimization is used to select the optimal nodes. Then the Gaussian Markov random field is used as the model for receiving signal in underwater distributed cooperative detection network, and the optimal node selection method based on information entropy is analyzed from the perspective of graph signal processing. Finally the computer simulation results for the selection of optimal nodes and reconstruction error of graph signal sampling are presented.