A Fog Computing-Based Collaborative Information Resource System for Smart Cities: TSAS and KLED Algorithms for Data Transmission and Service Deployment

Abstract

Effective integration and scheduling of information resources play a pivotal role in realizing intelligent management within the framework of smart city development. This research endeavors to overcome two significant hurdles: the redundancy in data transmission during the acquisition phase at the network's edge and the inefficiencies encountered when deploying analytical services across diverse edge devices. To address these challenges, a collaborative system architecture rooted in fog computing is introduced. A prediction mechanism, driven by spatiotemporal correlations, is incorporated to dynamically modulate data transmission intervals. This adjustment effectively curtails unnecessary synchronization, thereby enhancing the efficiency of data acquisition. Moreover, a deployment strategy based on multi-objective optimization is devised to allocate analytical tasks among edge devices constrained by limited resources, aiming to minimize the overall execution time. Experimental evaluations carried out on a real-world dataset encompassing 54 sensing terminals reveal that the proposed synchronization mechanism outperforms two traditional methods, reducing the false alarm rate by 58.90% and 31.35%, respectively, with a minimum mean absolute error of 2.6×10-5 . Additionally, the deployment strategy achieves an average reduction of 13.12% in service completion time across four standard scientific workflow structures. The system adeptly alleviates bandwidth constraints and computational limitations inherent in edge networks, providing a practical and effective solution for efficient data transmission and task scheduling in extensive smart city environments.