Comparative Study of Compressed Sensing for Heart Sound Acquisition in Wireless Body Sensor Networks

With the application of wireless body sensing network to real-time monitoring of biological signals, compressed sensing (CS) has become a promising signal acquisition technology that can extend monitoring time, reduce equipment cost and decrease power consumption. This paper makes a comparative study on the acquisition of heart sound (HS) signals by CS, compares and analyzes the performance of wavelet basis, reconstruction algorithms and frames. Among a large number of experimental data records, the reconstruction performance under different compression rates is comparatively studied, and the suitable wavelet basis, reconstruction algorithm and signal frame size for HS acquisition are obtained. The presented performance records and comparative analyses can give benchmarks for future academic research, and are also believed to be beneficial for practical applications.