Robust Federated Learning via Over-The-Air Computation

Abstract

This paper investigates the robustness of over-theair federated learning to Byzantine attacks. The simple averaging of the model updates via over-the-air computation makes the learning task vulnerable to random or intended modifications of the local model updates of some malicious clients. We propose a robust transmission and aggregation framework to such attacks while preserving the benefits of over-the-air computation for federated learning. For the proposed robust federated learning, the participating clients are randomly divided into groups and a transmission time slot is allocated to each group. The parameter server aggregates the results of the different groups using a robust aggregation technique and conveys the result to the clients for another training round. We also analyze the convergence of the proposed algorithm. Numerical simulations confirm the robustness of the proposed approach to Byzantine attacks.

Efficient Wireless Federated Learning with Partial Model Aggregation

“…According to the lower bound λ LB and upper bound λ UB , the optimal Lagrange multiply, λ * , can be solved by using the bisection search method. Furthermore, the optimal wireless bandwidth allocation policy θ t can be derived from (31). Based on the above analysis, we have the following remark.…”

“…However, the exact computation of the Lipschitz constant of deep learning architectures is intractable, even for two-layer neural networks [30]. Assumption 2 is widely used in the convergence analysis in FL algorithms, e.g., [14], [24], [31]. To begin with, we first derive a key lemma, proved in Appendix A, to assist our analysis as follows:…”

“…In (31), there still exists an unknown variable λ * . The value of λ * is determined by the equation K k=1 θ k,t (λ * ) = 1.…”