Power analysis of a lossless data compression technique for wireless wearable biometric devices

This paper presents a promising time-domain based lossless compression technique called Log2 Sub-band encoding, which is designed for using on wireless biomedical devices. Data compression can help to save power from the wireless transceiver during data transmission, and from the storage medium during reading and writing, ultimately leading to a longer battery life of the device. The performance of Log2 Sub-band is measured in terms of its compression ratio (CR) on EEG data and its power consumption. Our simulation results indicate a CR that is comparable and even superior to the well-known Huffman coding, whilst consuming minimal hardware resource. The simulations primarily use electroencephalogram (EEG) data, and the power consumption during compressing process is given to evaluate the system's improvement on its power performance. The possible influence of different biomedical data on technique's performance will also be considered and the signal classification potential of Log2 Sub-band will be noted.