SharkTooth: A Scalable Real-Time Algorithm for BLE-Based Wireless Body Sensor Networks Synchronization

The rapid expansion of Wireless Body Sensor Networks (WBSNs) in healthcare, rehabilitation, and movement analysis demands precise time synchronization across sensor nodes to ensure reliable multi-modal data fusion. Existing synchronization solutions for Bluetooth Low Energy (BLE)-based WBSNs often rely on hardware-level timestamping or operate in advertising mode, limiting their scalability, interoperability, or usability in real-time interactive applications. In this paper, we present SharkTooth, a novel, scalable, and real-time synchronization algorithm that operates entirely at the application layer of BLE. Unlike prior approaches, SharkTooth does not require any firmware modifications or specialized hardware, and it is compatible with commercial off-the-shelf BLE devices. The algorithm employs an adaptive timestamp correction technique that mitigates both deterministic drift and unpredictable delays caused by packet retransmissions and protocol variability. Extensive experimental validation using up to 12 sensor nodes and 18 BLE network configurations demonstrates submillisecond synchronization accuracy, with a median absolute average synchronization error between 0.21 ms and 0.62 ms, even in congested network scenarios. Compared to state-ofthe-art solutions, SharkTooth is uniquely capable of delivering robust, long-term synchronization in high-throughput, multinode systems using only connection-based BLE communication. By prioritizing compatibility, reproducibility, and real-world performance, SharkTooth advances the design of scalable and interoperable WBSNs for next-generation biomedical and wearable IoT applications.