A Five-Element USBL Location Method Based on DNN-Kalman Filter

Time Difference of Arrival (TDOA) is used for traditional Ultra-Short Baseline (USBL) positioning. Due to the short baseline length between acoustic arrays, the positioning accuracy is reduced in the case of low SNR. In order to solve this problem, this paper focuses on the phase information of underwater acoustic signal instead of using TDOA localization method. In this paper, five-element USBL array is used to filter underwater acoustic signals by means of long and short baseline multiplexing, deep neural network (DNN) and Kalman filtering algorithm. The simulation results show that the proposed method can effectively improve the positioning accuracy, and the positioning performance is much better than the traditional algorithm under the condition of low SNR.