EEG-Over-BLE: A Low-Latency, Reliable, and Low-Power Architecture for Multichannel EEG Monitoring Systems

Abstract

Recent enhancements in the field of wireless communication protocols have led to the development of several wearable devices and sensor networks suitable for health monitoring and diagnostic purposes. Nevertheless, new low-power solutions are still needed for enabling long-term, comfortable and lightweight monitoring systems. This paper deals with EEG-over-BLE, a novel architecture for multi-channel electroencephalography (EEG), characterized by low power consumption. The proposed architecture relies on a low-power analog front-end for EEG signals and a system-on-chip that supports the Bluetooth 5.0 protocol. Moreover, a near-lossless EEG compression algorithm and a simple channel encoding scheme are introduced in the proposed architecture in order to reduce energy consumption and improve network performance. Detailed investigations were carried out in order to evaluate the performance of the proposed architecture. In particular, simulation results and an analytical model on power consumption reported in this paper show that, when commercial small-form-factor batteries are used for power supply, the lifetime of a 64 channels EEG-over-BLE monitoring system is over 86 days. Moreover, the proposed architecture is highly reliable, achieving a packet loss rate lower than 0.1% and a latency below 35 ms, which is compliant with ISO/IEEE 11073 standard for medical device communications.