アナログ・ディジタル変換器出力の実時間圧縮

概要

論文の詳細を見る
We describe a fast lossless data compression algorithm suitable for digitized data taken at regular time intervals, such as outputs from analog-to-digital converters (ADCs). It is designed on the assumption that the present value can be predicted approximately from the past values, and that the distribution of the prediction error is approximately Gaussian with zero mean and small and slowly changing standard deviation. Unlike many offline compression tools such as LH and gzip, our algorithm does not need future values to encode the present value. This property is important for real-time transmission of compressed data on the network. The algorithm is to be integrated into our data acquisition system for the Large Helical Device (LHD) experiments at the National Institute for Fusion. Science (NIFS).
社団法人プラズマ・核融合学会の論文
1997-10-25