论文标题
无线联合学习与当地差异隐私
Wireless Federated Learning with Local Differential Privacy
论文作者
论文摘要
在本文中，我们在无线通道上研究了联合学习（FL）的问题，该通道由高斯多访问通道（MAC）建模，但受局部差异隐私（LDP）约束的约束。我们表明，无线通道的叠加性质为带宽有效梯度聚合提供了双重好处，并为用户提供了强大的LDP保证。我们提出了一种私人无线梯度聚合方案，该方案表明，当从$ k $用户中汇总梯度时，每个用户的隐私泄漏为$ \ nathcal {o} \ big（\ frac {1} {\ sqrt {k sqrt {k}}}} \ big）$与OrthogOnal Transpormation the Plusthenal eaperage aS a Ass a As A s an a As a As a Ass a As a Ass a a Ass a As a promphogonal tablese a salust s a s an a a As a As a Ass a As a Ass a As a Ass a As A instorta。我们还对拟议的私人FL聚合算法的收敛速率进行了分析，并研究了无线资源，收敛和隐私之间的权衡。
In this paper, we study the problem of federated learning (FL) over a wireless channel, modeled by a Gaussian multiple access channel (MAC), subject to local differential privacy (LDP) constraints. We show that the superposition nature of the wireless channel provides a dual benefit of bandwidth efficient gradient aggregation, in conjunction with strong LDP guarantees for the users. We propose a private wireless gradient aggregation scheme, which shows that when aggregating gradients from $K$ users, the privacy leakage per user scales as $\mathcal{O}\big(\frac{1}{\sqrt{K}} \big)$ compared to orthogonal transmission in which the privacy leakage scales as a constant. We also present analysis for the convergence rate of the proposed private FL aggregation algorithm and study the tradeoffs between wireless resources, convergence, and privacy.
