Content-extraction-based compression of acceleration data for mobile wireless sensors

A content-extraction based acceleration data compression algorithm is proposed for real-time transport quality monitoring in critical logistics applications. It splits the original data into three components, Tilt, Shock and Vibration, and compresses them separately by making use of the specific characteristics. Outstanding performances as well as low complexity and good scalability are achieved, which are essential for resource-constrained mobile wireless sensors. The algorithm has been proven by 46-day field test data. The compression ratio is up to 3.75 in loss-less mode, and up to 142 in lossy mode with 28.9dB SNR.