New research titled Minimising Data Loss in Bluetooth Low Energy (BLE): A Parallel Transmission and Application-Layer Modification Approach by Ahmad Asili investigates how BLE communication via the Web Bluetooth API can be made more reliable in industrial IoT (IIoT) settings.
BLE is widely used due to its low energy consumption, but it suffers from packet loss, limited throughput, and configuration constraints, especially when accessed via browsers using the Web Bluetooth API. The research sets out to explore how software-level solutions can improve BLE reliability without modifying hardware or the protocol stack.
Three experiments were conducted. The first tested whether using multiple BLE modules in parallel could reduce packet loss. Results showed that two parallel modules eliminated packet loss and minimised jitter, while three or four modules introduced diminishing returns or slight instability. The second experiment tested how different data transmission rates (1 Hz, 2 Hz, and 5 Hz) and packet sizes (50, 75, and 100 bytes) affected performance. A 2 Hz rate consistently gave the best results, balancing speed and reliability. The third experiment assessed long-term connection stability over five hours. With system-level improvements such as real-time scheduling and CPU pinning, jitter was reduced and packet reliability improved.
The research concluded that reliable and stable BLE communication using the Web Bluetooth API is achievable with careful application-level design. Key contributions include demonstrating that two BLE modules in parallel and a 2 Hz transmission interval optimise performance, and that software optimisations significantly improve long-term connection quality.