Data recoverability and estimation for perception layer in semantic web of things.

Area of Science:

• Computer Science
• Electrical Engineering
• Information Technology

Background:

• The Internet of Things (IoT) is rapidly evolving across various sectors, including industry, e-health, and education.
• Semantic Web of Things (SWoT) enhances IoT by enabling intelligent communication through a 7-layered architecture.
• The perception layer in SWoT is crucial for data collection but is prone to data loss due to resource limitations, link instability, and noise.

Purpose of the Study:

• To identify and address the challenges of data acquisition and loss in SWoT perception layers.
• To introduce a novel framework for data recovery and estimation in SWoT.
• To reduce data volume while maintaining accuracy.

Main Methods:

• A Compressive Sensing (CS) based framework is proposed for SWoT.
• The method leverages low-rank structures for accurate data recovery.
• Data estimation techniques are employed to reduce data volume.

Main Results:

• The proposed Compressive Sensing based Data Recoverability and Estimation (CS-RE) method demonstrates significant improvements.
• CS-RE outperforms existing reconstruction methods in terms of error ratio.
• The method achieves high accuracy in data recoverability on real sensory datasets.

Conclusions:

• The CS-RE method effectively addresses data loss issues in SWoT perception layers.
• The proposed framework offers a robust solution for reliable data acquisition and management in SWoT.
• This work contributes to enhancing the performance and reliability of SWoT systems.