车联网中云链融合隐私保护分布式计算卸载方案  

Cloud-Chain Fusion Privacy-Preserving Distributed Computation Offloading Scheme in Internet of Vehicles

作　　者：王强 赵全 王颖[1] 周福才[1] 徐剑[1] WANG Qiang;ZHAO Quan;WANG Ying;ZHOU Fu-Cai;XU jian(Software College,Northeastern University,Shenyang 110819;School of Cyber Science and Engineering,Xi’an Jiaotong University,Xi’an 710049)

机构地区：[1]东北大学软件学院,沈阳110819 [2]西安交通大学网络空间安全学院,西安710049

出　　处：《计算机学报》2025年第2期433-450,共18页Chinese Journal of Computers

基　　金：国家自然科学基金项目(62202090,62173101);辽宁省博士启动基金(2022-BS-077);中央高校基本科研业务费(N2417006);辽宁网络安全执法协同创新中心课题(XTCX2024-015)资助;by Liaoning Province Natural Science Foundation Medical-Engineering Cross Joint Fund under Grant 2022-YGJC-24;by the Fundamental Research Funds for the Central Universities under Grant N2217009.

摘　　要：随着车联网技术的迅猛发展,智能汽车中的车载应用越来越多地面临处理时延敏感且计算密集型任务的挑战。尽管车载计算单元具有一定的计算能力,但其有限的资源无法应对时延敏感的复杂任务。针对上述问题,计算卸载至资源丰富的云端是一种可行的解决方案。然而,远程云服务器的传输成本和通信延迟成为卸载时延敏感任务的主要瓶颈。随着移动边缘计算(MEC)兴起,计算能力从中心化的云端转移至网络边缘,降低了延迟,但现有方案仍存在容错性弱、隐私泄露和计算效率低等问题。本文针对这些不足,提出了一种云链融合的隐私保护分布式计算卸载方案(FCOS)。FCOS通过设计冗余分布式计算机制显著提升了容错性,有效减少了单点故障对计算的影响;利用区块链中的智能合约确保计算卸载过程的可验证性与公平性,增强了卸载的安全性;在隐私保护方面,结合同态加密和盲化因子技术,实现了在智能汽车端低计算开销下的数据隐私保护。与现有技术相比,FCOS在云计算阶段处理5000次以内多项式时,计算用时平均降低90.4%;在单点故障率为0%~60%时,10000次以内的多项式计算用时平均降低了92.9%~96.7%。理论分析与实验结果验证了该方案的高效性、安全性和可靠性。With the rapid development of Internet of Vehicles(IoV)technology,smart car applications are increasingly challenged by the need to handle latency-sensitive and computationintensive tasks.Although on-board computing units possess some processing capabilities,their limited resources often result in inefficiencies when dealing with such complex tasks.Offloading computations to resource-rich cloud servers is a viable solution,but the transmission cost and communication delays between the smart vehicle and remote cloud servers have become major bottlenecks for latency-sensitive tasks.Mobile Edge Computing(MEC)has emerged as a promising solution by shifting computing power from centralized clouds to the network edge,closer to the data source,which reduces latency.However,existing solutions still face challenges such as weak fault tolerance,privacy leaks,and low computational efficiency.To address these issues,this paper proposes a cloud-blockchain integrated privacy-preserving distributed computation offloading scheme(FCOS).FCOS significantly improves fault tolerance by designing a redundant distributed computing mechanism,which effectively reduces the impact of single-point failure on computation;it utilizes smart contracts in the blockchain to ensure the verifiability and fairness of the computation offloading process,which enhances the security of the offloading;and in terms of privacy protection,it combines the homomorphic encryption and the blinding factor technology to realize the data privacy protection under the low computation overhead at the smart car end.Compared with existing technologies,FCOS reduces the computing time by an average of 90.4%when processing polynomials within 5000 times in the cloud computing stage;when the single-point failure rate is 0%‒60%,the computing time of polynomials within 10000 times is reduced by an average of 92.9%‒96.7%.Theoretical analysis and experimental results verify the efficiency,security and reliability of the solution.

关 键 词：车联网 安全计算卸载 区块链 隐私保护 分布式计算 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]