HTO-UWSN: a hierarchical task offloading protocol for underwater wireless sensor networks in aquatic sports monitoring systems

Underwater sensor network (UWSN) has transformed aquatic sports like swimming, kayaking, and rowing through applications in sensing, motion analysis, and real-time feedback. However, the execution of complex and time-critical sports analytics tasks degrades its efficiency due to constrained hardware and energy resources. To overcome these constraints, this study proposes a hierarchical task offloading algorithm for underwater sensor networks (HTO-UWSN). The protocol is dynamic in nature and establishes a hierarchical cluster-based network architecture. It operates by deploying sensor nodes in the aquatic environment and wearable inertial sensor nodes attached to athletes. Nodes have been organized into a number of hierarchy levels in the form of a cluster headed by a resource-rich cluster head. The work of the leaf-level nodes includes performing on-body motion sensing and underwater data acquisition, while those at the intermediate nodes are for data forwarding and computation. When a task exceeds the processing or energy capacity of a lower-level node, it is offloaded to the nearest higher hierarchical level. The system uses a two-step hierarchical process. Firstly, ensemble learning classifies tasks for smart decision-making. Secondly, a graph partitioning algorithm is used to break complex tasks into parallel threads. Simulation results demonstrate that HTO-UWSN improves task execution rate by 67%, reliability by 55%, and scalability by 76%, suggesting its potential suitability for real-time aquatic sports monitoring systems.