计算机应用研究2026，Vol.43Issue(3)：712-719,8.DOI:10.19734/j.issn.1001-3695.2025.08.0268
面向智能交通的安全高效个性化联邦学习框架
Secure and efficient personalized federated learning framework for intelligent transportation
摘要
Abstract
Blockchain-federated learning(FL)integration enables decentralized and privacy-preserving collaboration in intel-ligent transportation systems,but device heterogeneity,non-IID data,and single-chain bottlenecks limit scalability and effi-ciency.This paper proposed a secure and cost-efficient personalized federated learning(SCPFL)framework.By leveraging blockchain sharding,SCPFL supported parallel training and aggregation across regions,while a cosine-similarity-based perso-nalization method balanced individual adaptation and cross-shard knowledge sharing.A dual mechanism of filtering and incen-tives further enhanced robustness and participation.Experiments on CIFAR-10,MNIST,and German traffic sign recognition benchmark(GTSRB)datasets show that SCPFL improves accuracy by up to 18.3％and accelerates convergence by 50％com-pared with the state-of-the-art personalized baselines(e.g.,Ditto).In terms of system efficiency,SCPFL achieves 53％higher throughput and reduces processing time,CPU,and memory usage by 33.3％,17.7％,and 19.7％over traditional main-subchain FL architectures.关键词
个性化联邦学习/区块链分片/安全/激励机制/智能交通系统Key words
personalized federated learning/sharding blockchain/security/incentive mechanism/intelligent transportation system(ITS)分类
信息技术与安全科学引用本文复制引用
王逸飞,胡颍,蔡婷,陈炜,吴雨芯,李晓丽,叶志伟..面向智能交通的安全高效个性化联邦学习框架[J].计算机应用研究,2026,43(3):712-719,8.基金项目
国家自然科学基金资助项目(62302154,U23A20318,62306108,62376089) （62302154,U23A20318,62306108,62376089）
湖北省自然科学基金资助项目(2024AFB882) （2024AFB882）
湖工大博士科研启动基金资助项目(XJ2022006701) （XJ2022006701）
智能感知系统与安全教育部重点实验室开放基金资助项目(KLISSS202404) （KLISSS202404）
湖北省教育厅科研重点项目(D20242602) （D20242602）
湖北省高等学校优秀中青年科技创新团队计划资助项目(T2023007) （T2023007）
