Non-cooperative target tracking method based on underwater acoustic sensor networks

Non-cooperative target tracking technology has significant applications in the field of ocean, serving both military and civilian purposes. The emerging underwater acoustic sensor networks (UASNs) have great potential for target tracking, but face challenges that include energy constraints, stratification effect, variable environment, and measurement outliers. To address these issues, a non-cooperative target tracking method based on UASNs is proposed in this paper. First, a mutual information (MI) function is established with the predicted target state and the received signal strength. Then, an MI-based active tracking node selection method is designed to reduce energy consumption and ensure tracking accuracy. After identifying active tracking nodes, a ray tracking method is used to calculate the distances under stratification effect. Subsequently, a robust and adaptive tracking algorithm is designed based on interactive multiple models assisted unscented Kalman filter to achieve accurate tracking, in which the equivalent noise covariance is utilized to change measurement noise adaptively, and a detector is implemented to detect and correct measurement outliers. The motion model transition probability is also adaptively adjusted based on the corrected data. Finally, the simulation and experimental results verify the feasibility and effectiveness of the proposed method.