Continuous Objects Tracking Based on Geometric Centroid of Feasible Region in USV-Assisted Underwater Acoustic Sensor Networks
Congpin Zhang,Li Liu,Yijia Wu,Zhengwei Xu,Changmao Wu
DOI: https://doi.org/10.1109/jiot.2023.3308516
IF: 10.6
2023-01-01
IEEE Internet of Things Journal
Abstract:Tracking continuous objects in the ocean, such as oil spills, is a significant challenge, and one that Underwater Acoustic Sensor Networks (UASNs) are well-suited to address. While prevalent approaches often employ complex, multi-phase processes to delineate the boundaries of these objects, enabling effective monitoring, they may not be ideally suited for time-sensitive situations, where a rapid response often outweighs the need for extreme precision. To address this gap, we introduce a streamlined boundary tracking algorithm, the Centroid Computation of Feasible Region for Continuous Object Tracking (CCFR-COT), intended for Unmanned Surface Vessel (USV)-assisted UASNs. This algorithm utilizes a fundamental full binary tree-structured network configuration along with two of its variations, designed to handle cases where continuous objects disperse in either free space or two semi-infinite spaces. The chosen network configuration aids in the selection of boundary mapping cells, thereby sketching a coarse-grained feasible region that envelops the actual boundary. Sequentially visiting these cells, the USV systematically constructs and deconstructs convex hulls of sensor nodes to progressively reduce location uncertainty within the feasible region. Only the centroid of the final feasible region in each boundary mapping cell is output for boundary fitting. Both experimental and simulated results validate the streamlined design of the CCFR-COT, which ensures timely tracking results and provides a balance between rapid response and tracking accuracy.
computer science, information systems,telecommunications,engineering, electrical & electronic