Hybrid Federated Learning for Noise-Robust Training
This work addresses noise robustness in distributed learning for privacy-enhanced applications, but it is incremental as it combines existing paradigms with adaptive optimizations.
The paper tackles the trade-offs between noise robustness and learning speed in federated learning and federated distillation by proposing a hybrid framework (HFL) that adaptively selects gradients or logits and optimizes weights and clustering. It shows that HFL achieves superior test accuracy at low SNR when exploiting degrees of freedom.
Federated learning (FL) and federated distillation (FD) are distributed learning paradigms that train UE models with enhanced privacy, each offering different trade-offs between noise robustness and learning speed. To mitigate their respective weaknesses, we propose a hybrid federated learning (HFL) framework in which each user equipment (UE) transmits either gradients or logits, and the base station (BS) selects the per-round weights of FL and FD updates. We derive convergence of HFL framework and introduce two methods to exploit degrees of freedom (DoF) in HFL, which are (i) adaptive UE clustering via Jenks optimization and (ii) adaptive weight selection via a damped Newton method. Numerical results show that HFL achieves superior test accuracy at low SNR when both DoF are exploited.