This works introduces a wireless signal processing platform based on a low-power FPGA. The architecture of the circuit is customized for the processing of sensor data used in biomedical applications such as ECG, EEG, or activity recognition using wearable computing. Hardware accelerators can be dynamically reconfigured to implement multiple signal processing tasks in a time-multiplexed manner. This approach allows reducing the size of the computing hardware while enabling energy-efficient operation. The functionality of the system is demonstrated for feature extractions in activity recognition applications.
展开▼
