Energy-Efficient Ubiquitous Sensor Collaboration System Enhancing Real-Time Decision Making in IoT Ecosystems

The rapid expansion of the Internet of Things (IoT) has necessitated energy-efficient, intelligent sensor collaboration for real-time decision-making. Existing models often fail to adapt to dynamic conditions or optimize energy use across distributed sensor networks. This study aims to address these limitations by proposing CE-STGNet (Collaborative Energy-aware Spatio-Temporal Graph Network), a novel framework that enhances decision accuracy while minimizing energy consumption. The model enables decentralized collaboration by modelling sensor interactions as a dynamic spatio-temporal graph, factoring in both node energy status and contextual importance. Utilizing the UNB CIC IoT 2023 dataset, which captures diverse real-time network behaviours, the framework was implemented using Python-based deep learning tools and evaluated across standard performance metrics. Results demonstrate CE-STGNet’s superior performance, achieving 99.96% accuracy, 99.95% precision, and 99.94% recall-surpassing state-of-the-art methods. Furthermore, it significantly reduced latency and energy usage, validating its efficiency in resource-constrained environments. In conclusion, CE-STGNet provides a scalable, intelligent approach to collaborative sensing, making it a promising advancement for sustainable, real-time IoT ecosystems.

Hali tarjima qilinmagan