摘要
Conventional federated learning (FL) approaches generally rely on a centralized server, and there has been a trend of designing asynchronous FL approaches for distributed applications partly to mitigate limitations associated with conventional (synchronous) FL approaches (e.g., single point of failure / attack). In this paper, we first introduce two new tools, namely: a quality-based aggregation method and an extended dynamic contribution broadcast encryption (DConBE). Building on these two new tools and local differential privacy, we then propose a privacy-preserving and reliable decentralized FL scheme, designed to support batch joining/leaving of clients while incurring minimal delay and achieving high model accuracy. In other words, our scheme seeks to ensure an optimal trade-off between model accuracy and data privacy, which is also demonstrated in our simulation results. For example, the results show that our aggregation method can effectively avoid low-quality updates in the sense that the scheme guarantees high model accuracy even in the presence of bad clients who may submit low-quality updates. In addition, our scheme incurs a lower loss and the extended DConBE only slightly affects the efficiency of our scheme. With the extended dynamic contribution broadcast encryption, our scheme can efficiently support batch joining/leaving of clients.
学科与关键词
- Privacy-Preserving and Reliable Federated LearningLu, Yi; Zhang, Lei*; Wang, Lulu; Gao, Yuanyuan21st International Conference on Algorithms and Architectures for Parallel Processing (ICA3PP), ELECTR NETWORK, 2021-12-03 To 2021-12-05.被引 1
- Aggregation Service for Federated Learning: An Efficient, Secure, and More Resilient RealizationZheng, Yifeng*; Lai, Shangqi; Liu, Yi; Yuan, Xingliang; Yi, Xun; Wang, Cong*IEEE Transactions on Dependable and Secure Computing, 2023, 20(2): 988-1001.被引 22
- Privacy-Preserving Federated Learning via Functional Encryption, RevisitedChang, Yansong; Zhang, Kai; Gong, Junqing*; Qian, HaifengIEEE Transactions on Information Forensics and Security, 2023, 18: 1855-1869.被引 1
