用神经网络实现隐私保护联邦学习，解决密钥分发难题
Privacy-Preserving Federated Learning via Homomorphic Adversarial Networks
- 用同态对抗网络模拟多密钥加密，避免密钥共享
- 准确率损失≤1.35%，比传统方法快6075倍
- 适合对隐私和效率要求高的联邦学习场景
隐私保护联邦学习（PPFL）旨在训练全局模型的同时保护各参与方的数据隐私。现有协议存在准确率下降、需共享密钥或协作生成/解密密钥等问题。为此，我们提出首个基于神经网络的PPFL协议，结合专为PPFL设计的可聚合混合加密方案。所提出的同态对抗网络（HANs）证明神经网络可实现类似多密钥同态加密（MK-HE）的功能，同时解决密钥分发与协同解密问题。实验表明，HANs对隐私攻击具有鲁棒性。在多个数据集上，其准确率损失不超过1.35%，相比非私有联邦学习几乎无损；相较于传统MK-HE方案，加密聚合速度提升6,075倍，通信开销增加29.2倍。
原文摘要 · Abstract (English)
Privacy-preserving federated learning (PPFL) aims to train a global model for multiple clients while maintaining their data privacy. However, current PPFL protocols exhibit one or more of the following insufficiencies: considerable degradation in accuracy, the requirement for sharing keys, and cooperation during the key generation or decryption processes. As a mitigation, we develop the first protocol that utilizes neural networks to implement PPFL, as well as incorporating an Aggregatable Hybrid Encryption scheme tailored to the needs of PPFL. We name these networks as Homomorphic Adversarial Networks (HANs) which demonstrate that neural networks are capable of performing tasks similar to multi-key homomorphic encryption (MK-HE) while solving the problems of key distribution and collaborative decryption. Our experiments show that HANs are robust against privacy attacks. Compared with non-private federated learning, experiments conducted on multiple datasets demonstrate that HANs exhibit a negligible accuracy loss (at most 1.35%). Compared to traditional MK-HE schemes, HANs increase encryption aggregation speed by 6,075 times while incurring a 29.2 times increase in communication overhead.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。