.png){kind=icon}
.png){kind=icon}
Return
ACTF: An efficient lossless compression algorithm for time series floating point data
.png){kind=icon}
OA
AI
DOI：10.1016/j.jksuci.2024.102246.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The volume of time series data across various fields is steadily increasing. However, this unprocessed massive data challenges transmission efficiency, computational arithmetic, and storage capacity. Therefore, the compression of time series data is essential for improving transmission, computation, and storage. Currently, improving time series floating-point coding rules is the primary method for enhancing compression algorithms efficiency and ratio. This paper presents an efficient lossless compression algorithm for time series floating point data, designed based on existing compression algorithms. We employ three optimization strategies data preprocessing, coding category expansion, and feature refinement representation to enhance the compression ratio and efficiency of compressing time-series floating-point numbers. Through experimental comparisons and validations, we demonstrate that our algorithm outperforms Chimp, Chimp128, Gorilla, and other compression algorithms across multiple datasets. The experimental results on 30 datasets show that our algorithm improves the compression ratio of time series algorithms by an average of 12.25% and compression and decompression efficiencies by an average of 27.21%. Notably, it achieves a 24.06% compression ratio improvement on the IOT1 dataset and a 42.96% compression and decompression efficiency improvement on the IOT4 dataset.
Keywords:
Time series floating point data
Lossless compression
Internet of things
Compression algorithm
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.