Asynchronous tracking of sensor nodes in an internet of underwater thing (IoUT) system

The technology of the Internet of Underwater Things (IoUTs) is a novel concept for underwater applications that faces several challenges, including asynchronous and mobile sensor nodes with unknown positions over time. Online tracking of these asynchronous IoUT components is a critical aspect of functional IoUT systems. It requires minimal computation due to the limited battery capacity of underwater sensor nodes, as well as accurate values for anchor node clocks and sound propagation speed parameters. Extended Kalman Filter (EKF)-based tracking is a method that enables online tracking. However, the challenge lies in how the EKF can be effectively utilized given these constraints. Due to stratification effects, underwater sound propagation speed varies with depth, introducing unknown parameters that must be estimated. In real-world scenarios, even anchor nodes with known locations can be asynchronous. While other algorithms may not account for these factors, this paper presents a method that combines an EKF-based tracking approach with a Weighted Least Squares (WLS) synchronization method. To achieve this, we employ a technique that generates new measurements from raw data in a manner suitable for EKF application. These new measurements are independent of the sensor node’s clock parameters but depend on the clock parameters of the anchor nodes. To perform tracking, we first use linear and nonlinear methods to estimate the clock skew, clock offset, and sound velocity parameters of the anchor nodes. Then, we incorporate the estimated clock parameters of the anchor nodes into the new measurements. Finally, we apply the EKF to these modified measurements. The results presented in this work are obtained using the estimated clock parameters of the anchor nodes and the sound speed profile parameters. Therefore, this study also examines the impact of anchor node synchronization and the estimation of a linearly depth-dependent sound propagation speed on the proposed sensor node tracking and time synchronization method. The proposed tracking algorithm is based on the EKF method, followed by the estimation of the sensor node's clock parameters using the WLS algorithm. Simulation results validate the theoretical analysis, and the findings are compared with other methods and the Cramér-Rao Bound to confirm the superiority and effectiveness of the proposed approach.