Design of a Low-Cost Wireless Body Area Network for Real-Time Electromyography and Inertial Data Acquisition

Wireless Body Area Networks (WBANs) for surface electromyography (EMG) and inertial measurement are essential tools in rehabilitation, enabling objective assessment of muscle activity and movement during therapy sessions. The advancement of wearable technology also offers significant potential to transform rehabilitation by enabling continuous, objective monitoring of patient progress outside of clinical settings. However, the adoption of such technology is often hindered by the limitations of commercial devices, which can be prohibitively expensive, operate as closed systems with limited data access, and lack the flexibility to be adapted for specific research needs. Furthermore, challenges related to power consumption, which limits recording duration, and the difficulty of synchronizing data from multiple sensors across the body remain significant barriers to capturing accurate, coordinated movement and muscle activity. This thesis presents the design, implementation, and validation of a low-cost, modular WBAN capable of simultaneously recording high-fidelity EMG signals and inertial measurement unit (IMU) data from multiple body locations. This work addresses several critical challenges, including power optimization for battery-operated wearables, multi-node time synchronization, real-time constraints, and the attainment of signal quality comparable to that of research-grade instrumentation. Experimental validation confirms the effectiveness of the proposed system. The measured noise floor remains within the low microvolt range, resulting in a high signal-to-noise ratio consistent with reliable EMG acquisition. Time synchronization across distributed nodes achieves sub-millisecond accuracy under typical operating conditions, ensuring precise temporal alignment of multimodal data. Furthermore, the system demonstrates stable continuous operation exceeding three hours on compact lithium-polymer batteries, while sustaining sampling frequencies of 2kHz for EMG acquisition. Results demonstrate that modern embedded platforms can achieve signal quality and synchronization performance suitable for rehabilitation monitoring while maintaining system cost substantially below commercial alternatives, improving accessibility for research laboratories and clinical rehabilitation centers with limited budgets.