Underwater multitarget fusion tracking method for passive sonar

The marine environmental noise will cause the detection results of weak targets significantly different in subdbands and lead to the performance degradation problem of tracking algorithms based on fullband detection results. To address this issue, a subband fusion tracking method is proposed. A modified Gaussian mixture probability hypothesis density (GM-PHD) filter is introduced to obtain direction of arrival (DOA) tracking results for different subbands. In addition, the subband tracking results are fused by the generalized covariance intersection (GCI) technique to obtain the tracking results with integrated subband information. Simulation results illustrate that the proposed method can improve the tracking ability of weak targets with different signal-to-noise ratios in each subband, and the computation time is relatively close to the comparison methods. The sea trial data processing results further demonstrate the effectiveness of the proposed method.