Asynchronous Binary Compressive Sensing for Wireless Body Sensor Networks

Next-generation Wireless Body Sensor Networks (WBSNs) calls for miniaturization and power-efficient integration for long-term monitoring, real-time diagnostics and patient-centered healthcare solutions. However, state-of-the-art WBSN prototypes remain challenged by stringent power constraints and large form factors. The recently proposed asynchronous compressive sensing scheme suggests an efficient way to improve power consumption by reducing the data volume in energy-hungry radio links. In this paper, we present a modified front-end called Asynchronous Binary Compressive Sensing (ABCS) for WBSNs. A low-cost reconstruction method is proposed that exploits the embedded binary signal structure in the ABCS. By incorporating binary amplitude as a prior, better signal recovery performance is obtained comparing with the traditional approaches. Analyses and simulations with an ECG recording confirm the ABCS front-end outperforms the conventional CS approaches in terms of hardware complexity, power consumption and system flexibility.