A Game Theoretical Priority-Aware R2V Task Offloading Framework for Vehicular Fog Networks

Abstract

Modern vehicles, equipped with advanced digital infrastructure, are transforming from mere transportation units to powerful mobile servers. This trend, coupled with the rise of mobile edge computing (MEC), presents a unique opportunity to leverage vehicles as temporary computing resources. Roadside Units (RSUs) can then offload urgent tasks to passing vehicles, enabling ubiquitous connectivity and low-latency services. However, efficient task offloading remains a challenge. This paper proposes a two-level resource management framework for Priority-Aware RSU-to-Vehicle (PA-R2V) Task Offloading (PA-R2VTO) to address this issue. The first level utilizes a two-stage Stackelberg game to incentivize vehicles within RSU coverage to allocate a portion of their computing resources. The second level focuses on task assignment to participating vehicles. Two computation modes are considered: serial and parallel. Serial computation leverages a Multiple Knapsack Problem (MKP) formulation to prioritize tasks based on deadlines, while parallel computation employs a non-cooperative congestion game model. Both approaches are compared against the traditional Earliest Deadline First (EDF) scheduling algorithm. Extensive simulations validate the framework’s effectiveness, paving the way for a more tangible realization of the Vehicle-as-a-Resource (VaaR) concept.