Quantum-based predictive fog scheduler for IoT applications

Abstract

Load scheduling across distributed fog computing nodes has been a major challenge to meet the increased demand of real-time data analysis, and time-sensitive decision-making. This study presents a Quantum Computing-inspired (QCi) optimized load scheduling technique in fog computing environments for real-time IoT applications. In addition to this, QCi-Neural Network Model is used as a predictive model to determine the optimal computational node for enabling real-time service delivery. For validation, simulations were performed using 3, 5, 7, and 10 fog nodes at different instances to schedule nearly 600 heterogeneous tasks to obtain the respective results. Results were compared with other state-of-the-art scheduling models like Min–Max, Minimum Completion time, and Round Robin. Better results were registered for the proposed technique in terms of minimum average task completion time (28.15 s for 10 nodes) and average energy consumption (3.48 J for 10 nodes). Moreover, higher rates of statistical parameters like sensitivity (88.66%), specificity (91.28%), precision (92.25%), and coverage (95.66%) were acquired depicting enhanced overall performance.