Mx-TORU: Location-aware multi-hop task offloading and resource optimization protocol for connected vehicle networks

Abstract

Connected Vehicle Networks (CVNs), as a part of Internet of Vehicles (IoV), represent an innovative solution for enhancing communication between vehicles and Internet of Things (IoT) devices within transportation infrastructures. However, task offloading in CVNs presents significant challenges due to high computational demands and the dynamic nature of network conditions. While traditional static fog networks support CVNs, they often suffer from inefficiencies in resource allocation, leading to underutilization or over-utilization, as well as elevated maintenance costs. To address these limitations, mobile fog computing emerges as a more adaptable solution, enabling efficient task processing by leveraging the resources of nearby vehicles. In this paper, we introduce a novel mobility-driven protocol, Mx-TORU, which combines multi-hop task offloading with resource optimization to enhance task processing efficiency in CVNs. This protocol builds upon our previously proposed MobTORU framework, aiming to maximize resource utilization through dynamic multi-hop strategies. Extensive experiments using real-world vehicular mobility data demonstrate that Mx-TORU improves resource utilization by up to 17.8% compared to one-hop methods. Additionally, our Mx-TORU protocol and the employed RELiOff algorithm show a consistent improvement of at least 5% in task offloading efficiency across various test scenarios including intelligent transformation systems.