Joint task offloading and resource allocation scheme with UAV assistance in vehicle edge computing networks

Abstract

As an emerging and promising technology paradigm, Vehicle Edge Computing (VEC) aims to enhance the performance and user experience of in-vehicle applications through efficient computation offloading strategies. However, with the increasing demand for high-complexity, computationally intensive applications within the automotive industry, VEC systems are facing the challenge of limited resources, and how to effectively manage and utilize the limited computational resources has become an urgent problem. In this paper, we propose a novel framework for UAV-assisted task offloading and resource allocation in VEC networks. The framework integrates Software Defined Networking (SDN) and Unmanned Aerial Vehicles (UAVs) to improve computation efficiency and resource utilization. The utility functions of requesting vehicle and VEC server are defined, and an incentive mechanism is proposed to encourage multiple UAVs to form an effective resource pool that can be used for VEC tasks. Based on the Stackelberg game to optimize task offloading and resource allocation and ensure well collaboration among VEC servers, UAVs, and vehicles, the existence of a Nash equilibrium is proved by theoretical derivation. Subsequently, we adopt an efficient evolutionary strategy-genetic algorithm to explore and optimize the optimal pricing strategy for VEC servers. Also, a task allocation algorithm is designed and implemented, which aims to maximize the revenue of UAVs by minimizing the cost of UAV coalition. Finally the simulation comparison experiments are conducted, and the results strongly validate the effectiveness and feasibility of the proposed scheme.