Study of low power wireless protocols for event-driven sEMG acquisition systems

Nowadays smart healthcare constitutes an active field of research that offers the potential of great improvement in the delivery and monitoring of healthcare. The expansion of this sector is supported by the recent development of sensors and embedded systems, wearable and easily integrated into everyday life. New higher performances wireless technologies have empowered the development of Wireless Body Area Networks (WBANs) as the link between different devices placed on the human body offering constant monitoring of relevant parameters without posing constraints on normal daily life activities. This master thesis project aims to analyse and characterize performances and peculiarity of different wireless protocols to manage the event drive data transfer from an acquisition board to a host station, in which these will be processed. The real-time transmission of digital events generated from the acquisition of biomedical signals like sEMG (surface Electromyography) was investigated with the Average Threshold Crossing (ATC) technique. This methodology allows a low power acquisition due to a high reduction in the data size, and to the entire circuitry complexity. The ATC event-driven technique in facts registers for a new event each time the signal overcomes a threshold. The hardware selected for the implementation is a Multi-Protocol System-on-Chip (SoC) with low power consumption to maximize application energy efficiency and battery life. It is a highly flexible single-chip solution for short-range wireless applications adopting one between Bluetooth Low Energy (BLE), ANT, Bluetooth Mesh, Bluetooth 5.0, IEEE 802.15.4, Thread or ZigBee protocol stack to transfer packets among the devices in the network. The setup consists of a specific WBAN composed by one client’s and one server’s board to test each wireless protocol. The performances of the network have been assessed in terms of average current consumption, latency and event rate between the delivered and received a packet. About current consumption, BLE and ANT showed extremely good results with an average current peak of about 1mA. Taking into account the latency ANT and Thread have values around 5 ms for the higher data rate even if the second one has a standard deviation of that measure really close to zero. Considering the event data rate transmission promising result have been reached with BLE and Thread with at least 200 events per second. From an overall perspective, the most promising protocols for an event-driven sEMG acquisition system are BLE for the extremely low energy demand and Thread for the effective latency of few milliseconds at the upper data rate.