.png){kind=icon}
.png){kind=icon}
Return
Waved zero streaming: A time series compression framework with adaptive segmentation optimization
.png){kind=icon}
PRE
AI
DOI：10.1016/j.ipm.2026.104697.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The proliferation of time-series data from Internet of Things devices has made efficient compression mechanisms critical for large-scale data management. However, existing state-of-the-art algorithms often flounder when facing dynamic data fluctuations due to their reliance on rigid encoding strategies, while general-purpose compressors incur prohibitive latency. To bridge this gap, this paper introduces Wave Zero Streaming (WZS), a novel lossless compression framework specifically architected for non-stationary data streams. WZS synergizes three pivotal innovations: adaptive segmentation to isolate local fluctuation intensity, a heuristic genetic algorithm-based optimizer to fine-tune compression parameters dynamically, and a stream processing module enabling real-time adaptive algorithm selection. Extensive empirical evaluations across 43 diverse datasets reveal that WZS outperforms leading algorithms–including Camel, Chimp, and ACTF–achieving an average compression ratio improvement of 36.41% (peak: 53.08%), with a notable 13.28% enhancement specifically over the state-of-the-art Camel algorithm. Moreover, WZS surpasses general compression methods by 15.07% in compression ratio while reducing computational latency by two orders of magnitude, and exhibits a 16.45% speed advantage over comparable floating-point streaming baselines. These findings validate WZS as a scalable, high-performance solution that effectively balances efficiency and speed for resource-constrained IoT and edge computing environments.
Keywords:
Time-series compression
Adaptive segmentation
Streaming data
Lossless compression
IoT edge computing
Journal
I
IF:
6.9
Papers:
310
Citations:
0