Deadline-Aware Adaptive Fuzzification for Task Offloading in Vehicle-to-Vehicle Networks

Task offloading in Vehicle-to-Vehicle (V2V) networks distributes computational workloads to nearby vehicles, thereby enhancing resource utilization and minimizing latency. This process is complex due to uncertainties, such as variable network conditions and mobility, where applying fuzzy logic (fuzzification) offers a robust solution to manage such unpredictability. However, existing methods often overlook strict deadlines for tasks and dynamic vehicular contexts, leading to suboptimal performance. This paper proposes the Deadline-Aware Adaptive Fuzzification (DAAF) approach to address these challenges in V2V communications within Vehicle-to-Everything (V2X) systems. Motivated by the need to select adaptive fuzzy rules and deadline-sensitive offloading in highly dynamic environments, DAAF integrates the Computational Power Factor (CPF), Mobility Factor (MF), Link Quality Factor (LQF), and task deadlines into its fuzzy logic framework, ensuring timely and efficient task execution. Simulations using Omnet++, SUMO, and Veins show that the DAAF approach reduces the Average Completion Time (ACT) by 12% at 250 vehicles, achieves a peak Task Completion Ratio (TCR) of 90% at 150 vehicles with a 15% improvement over static fuzzy logic, and enhances the Average Execution Time (AET) by 7.7%, outperforming baseline methods in dense urban scenarios.