Design and Implementation of a Compressed Sensing Encoder: Application to EMG and ECG Wireless Biosensors

Abstract

Among the existing applications of wireless body sensor networks (WBSNs), a wearable health monitoring system (WHMS) is the most important. In a typical WHMS, miniature wireless biosensors, attached to or implanted in the human body, collect bio-signals such as the electrocardiogram (ECG), blood pressure or electromyogram (EMG) to provide real time and continuous health monitoring. In this paper, we present a compressed sensing (CS)-based approach to compress and recover the sensed bio-signals from the wireless biosensors of a WBSN. The CS encoding process has a low computational complexity and is suitable for use in power-constrained systems such as WHMS. We propose a simple deterministic measurement matrix, which is easy to implement in hardware. We design a digital CS encoder implementing the proposed measurement matrix and use it to compress the bio-signals in EMG and ECG wireless biosensors. The simulations and experimental results have shown that the EMG and ECG signals are compressed and recovered without perceptible loss if the compression ratios are, respectively, less than or equal to 75 and 87.5%. The obtained results have also confirmed the simplicity of the proposed measurement matrix since the CS encoder does not affect the memory usage or the processing time of the microcontrollers embedded in the wireless biosensors. Additionally, the CS encoder decreases by up to 75 and 87.5% the energy consumption of the transceivers for the EMG and ECG wireless biosensors.