摘要
Federated learning(FL)is a distributed machine learning paradigm that excels at preserving data privacy when using data from multiple parties.When combined with Fog Computing,FL offers enhanced capabilities for machine learning applications in the Internet of Things(IoT).However,implementing FL across large-scale distributed fog networks presents significant challenges in maintaining privacy,preventing collusion attacks,and ensuring robust data aggregation.To address these challenges,we propose an Efficient Privacy-preserving and Robust Federated Learning(EPRFL)scheme for fog computing scenarios.Specifically,we first propose an efficient secure aggregation strategy based on the improved threshold homomorphic encryption algorithm,which is not only resistant to model inference and collusion attacks,but also robust to fog node dropping.Then,we design a dynamic gradient filtering method based on cosine similarity to further reduce the communication overhead.To minimize training delays,we develop a dynamic task scheduling strategy based on comprehensive score.Theoretical analysis demonstrates that EPRFL offers robust security and low latency.Extensive experimental results indicate that EPRFL outperforms similar strategies in terms of privacy preserving,model performance,and resource efficiency.
基金
supported in part by the National Natural Science Foundation of China(62462053)
the Science and Technology Foundation of Qinghai Province(2023-ZJ-731)
the Open Project of the Qinghai Provincial Key Laboratory of Restoration Ecology in Cold Area(2023-KF-12)
the Open Research Fund of Guangdong Key Laboratory of Blockchain Security,Guangzhou University。
作者简介
Ke Zhijie received the B.S.degree in data science and big data technology from Kunming University of Science and Technology(KUST),Kunming,China,in 2022.He is currently pursuing an M.S.degree in artificial intelligence from Qinghai University.His current research is federated learning and data security;corresponding author:Xie Yong received the Ph.D degree in computer science from Wuhan University,Wuhan,China,in 2016.He is currently a Professor with School of Computer and Information Science,Qinghai institute of Technology,China.His current research interests include network and information security,artificial intelligence applications and security,and industrial digitization,email:mark.y.xie@qq.com;Syed Hamad Shirazi is currently working as an Assistant Professor with the Department of Computer Science and Information Technology,Hazara University Mansehra,Pakistan.He is also the head of department.His research interests include computer vision,texture analysis,neural networks,object recognition,pattern recognition,digital image processing,machine learning,wavelet transformation and artificial intelligence security.He is also working as Co-PI on HEC funded Project“Deep Learning Based Prediction Support System for Anaemic RBCs”in collaboration with Shaukat Khanam Cancer Hospital Lahore Pakistan.He has more than 10 years of experience in medical imaging.He published more than 35 research articles in high impact factor journals;Li Haifeng received the Ph.D degree in computer science from the School of Software,Dalian University of Technology,Dalian,China.He is currently an Associate Professor with the School of Computer and Information Science,Qinghai Univiersity of Science and Technology.His research interests include applied cryptography,information security,and artificial intelligence.
