西安电子科技大学学报（自然科学版）2025，Vol.52Issue(3)：85-98,14.DOI:10.19665/j.issn1001-2400.20250301
区块链赋能的车辆边缘网络任务卸载方法研究
Research on the task offloading methodology for vehicular edge networks enabled by the blockchain
摘要
Abstract
With the booming development of urban intelligent transportation and the Internet of Vehicles(IoV),the demand for computing and cache resources has been increasing thanks to computationally intensive applications such as autonomous driving,image and voice processing.Considering that in the traditional cloud computing architecture of the Internet of Vehicles,the cloud servers are located in the core of the network,resulting in large propagation delays and making it difficult to provide real-time and high-quality services for moving vehicles.This study aims to propose a research on the task offloading methodology for vehicular edge networks enabled by the blockchain(BTO-VEN).Through task offloading,the computing tasks generated by vehicles are dynamically allocated to edge computing servers or neighboring vehicles,thereby reducing the computational load and latency.First,edge computing(MEC)is integrated into the Internet of Vehicles to form a vehicular edge network(VEN),and MEC servers are deployed at the network edge to reduce data communication latency and improve the computing efficiency of task offloading.Then,a dual deep learning algorithm of prediction-reward-detection(PRD-DDQN)is designed,which combines deep neural networks and the Q-learning algorithm to carry out strategy training and learning.The PRD training process is used to estimate the cumulative rewards between two adjacent states and incorporate these prediction results into the algorithm training to achieve dynamic decision-making to adapt to changing environments.Finally,blockchain technology is integrated into the MEC servers,and the integrity and reliability of task offloading transactions are maintained through a private blockchain structure.Experimental results show that compared with the traditional MADDPG method,the proposed method has great advantages,with a 39.98％increase in cumulative rewards and a 20.09％reduction in energy consumption.关键词
区块链/边缘计算/深度学习/车联网/任务卸载Key words
blockchain/edge computing/deep learning/vehicular edge computing/task offloading分类
信息技术与安全科学引用本文复制引用
康海燕,刘鑫旭,李彦芳..区块链赋能的车辆边缘网络任务卸载方法研究[J].西安电子科技大学学报（自然科学版）,2025,52(3):85-98,14.基金项目
国家社会科学基金(21BTQ079) （21BTQ079）
未来区块链与隐私计算高精尖创新中心基金(GJJ-23) （GJJ-23）
北京市教育委员会科技计划(KM202011232022) （KM202011232022）
