区块链赋能联邦学习:方法、挑战与展望

Blockchain Enabled Federal Learning: Approaches, Challenges, and Prospects

摘要: 针对区块链技术与联邦学习(federated learning,FL)结合后在安全、隐私等方面存在的问题,对区块链赋能FL中的相关方法进行综述与分析。首先,分别阐述了FL和区块链的基础知识,并在此基础上总结了区块链赋能FL的前沿通用架构;其次,研究了目前安全、隐私、激励以及效率方法的进展,分析了相关文献之间的联系和各自的优缺点;最后,指出区块链赋能FL目前存在的问题,提出解决方案并进行了展望。

Abstract: In response to security and privacy in the integration of blockchain technology with federated learning (FL), a comprehensive review and analysis of the relevant methods for empowering FL with blockchain are provided.First, the fundamental knowledge of FL and blockchain is elucidated separately, and on the basis of this, the state-of-the-art general architectures for blockchain-enabled FL are summarized.Second, the progress in security, privacy, incentives, and efficiency methods is investigated and the connections between relevant literature and their respective advantages and disadvantages are analyzed.Finally, the current issues in blockchain-enabled FL are identified, and potential solutions are proposed, along with future prospects.