TG-SPRED: Temporal graph for sensorial data PREDiction

Laidi, Roufaida; Djenouri, Djamel; Djenouri, Youcef; Lin, Jerry Chun-Wei

Abstract

This study introduces an innovative method aimed at reducing energy consumption in sensor networks by predicting sensor data, thereby extending the network's operational lifespan. Our model, Temporal Graph Sensor Prediction (TG-SPRED), predicts readings for a subset of sensors designated to enter sleep mode in each time slot, based on a non-scheduling-dependent approach. This flexibility allows for extended sensor inactivity periods without compromising data accuracy. TG-SPRED addresses the complexities of event-based sensing - a domain that has been somewhat overlooked in existing literature - by recognizing and leveraging the inherent temporal and spatial correlations among events. It combines the strengths of Gated Recurrent Units and Graph Convolutional Networks to analyze temporal data and spatial relationships within the sensor network graph, where connections are defined by sensor proximities. An adversarial training mechanism, featuring a critic network employing the Wasserstein distance for performance measurement, further refines the predictive accuracy. Comparative analysis against six leading solutions using four critical metrics - F-score, energy consumption, network lifetime, and computational efficiency - showcases our approach's superior performance in both accuracy and energy efficiency.

Citation

Laidi, R., Djenouri, D., Djenouri, Y., & Lin, J. C. (2024). TG-SPRED: Temporal graph for sensorial data PREDiction. ACM Transactions on Sensor Networks, 20(3), 1-20. https://doi.org/10.1145/3649892

• Journal Article Type: Article
• Acceptance Date: Feb 7, 2024
• Online Publication Date: Feb 28, 2024
• Publication Date: Apr 13, 2024
• Deposit Date: Mar 1, 2024
• Publicly Available Date: Mar 29, 2024
• Journal: ACM Transactions on Sensor Networks
• Print ISSN: 1550-4859
• Electronic ISSN: 1550-4867
• Publisher: Association for Computing Machinery (ACM)
• Peer Reviewed: Peer Reviewed
• Volume: 20
• Issue: 3
• Article Number: 64
• Pages: 1-20
• DOI: https://doi.org/10.1145/3649892
• Keywords: CCS Concepts: • Computer systems organization ! Sensor networks; Embedded systems; • Hardware ! Power and energy Additional Key Words and Phrases: graph convolution neural network, adversarial training, sensor energy savings, spatiotemporal learning
• Public URL: https://uwe-repository.worktribe.com/output/11753262