Lightweight Time-Series Signal Compression Period Extraction and Multiresolution using Difference Sequences

Research output: Contribution to journal/Conference contribution in journal/Contribution to newspaperJournal articleResearchpeer-review

2 Citations (Scopus)

Abstract

In the Internet of Things (IoT), connected devices generate a massive amount of data that need to be processed and transmitted to the data aggregator or edge device. The connected devices are resource-constrained in terms of memory, computation power, and energy. In this paper, we propose a novel transform using difference sequences. The proposed transform is multiplierless, which makes it very promising for resource-constrained IoT devices. Various properties of the difference sequences, such as orthogonality, linear independence, and circular shift, are studied in detail. These sequences are sparse and take values from the set {0,1,-1}, which make these sequences very efficient in computation. Applications of the proposed transform are shown for lossless compression, period extraction, and multiresolution using electrocardiogram, accelerometer, images, and photoplethysmography datasets. Furthermore, the proposed transform is compared with the state-of-the-art data compression transforms.
Original languageEnglish
JournalIEEE Internet of Things Journal
Volume9
Issue9
Pages (from-to)7043-7050
Number of pages8
DOIs
Publication statusPublished - May 2022

Keywords

  • Discrete derivatives
  • Internet of Things (IoT)
  • Lossless compression
  • Multiresolution
  • Period estimation
  • Ramanujan sums
  • Resource-constrained

Fingerprint

Dive into the research topics of 'Lightweight Time-Series Signal Compression Period Extraction and Multiresolution using Difference Sequences'. Together they form a unique fingerprint.

Cite this