Publication

Abstract : Persistent health monitoring is the key feature in present day wearable health monitoring system. The focus is on reducing the power consumption associated with transmission of large data content by reducing the bandwidth required. Signals sampled wastefully at Nyquist rate increases power dissipation drastically when RF Power amplifier (inside the body area networks of wearable device) transmits sensed data to personal base station. Compressed Sensing (CS) is an emerging technique that condenses the information in the signal into a lower dimensional information preserving domain before sampling process. CS facilitates data acquisition at sub-Nyquist frequencies. The original signal is reconstructed from the compressively sampled signal by solving an undetermined system of linear equations. In this paper a scalable hardware for CS in ECG signal is modeled. The factors determining the quality of reconstruction of a Compressively Sampled ECG signal is studied in both time and wavelet domain using the modeled hardware.