Less is more. The limited memory and battery power of wearables necessitates rapid and efficient transmission of data. Reducing the size of transmitted data via compression algorithms, including run-length encoding and Huffman encoding, is one way to minimize power consumption and facilitate storage of ever-accumulating amounts of data.
This lossy method of wearable data compression involves six major steps — the discrete wavelet transform, thresholding, scaling, quantization, combination, and compression — and was tested on five types of data — accelerometry (ACC), electrocardiogram (ECG), electrodermal activity (EDA), photoplethysmogram (PPG), and temperature (TEMP).
BME+ECE at Duke | in search of what to do with his talents