Joint Superposition Coding and Training for Federated Learning over Multi-Width Neural Networks

TL;DR

This paper introduces SlimFL, a novel federated learning framework that combines superposition coding and training with width-adjustable neural networks to enhance communication efficiency and robustness over wireless channels.

Contribution

It proposes a new FL method integrating superposition coding and training for multi-width neural networks, addressing data heterogeneity and channel variability.

Findings

SlimFL improves communication efficiency in federated learning.

It effectively handles non-IID data distributions.

The approach is robust under varying wireless channel conditions.

Abstract

This paper aims to integrate two synergetic technologies, federated learning (FL) and width-adjustable slimmable neural network (SNN) architectures. FL preserves data privacy by exchanging the locally trained models of mobile devices. By adopting SNNs as local models, FL can flexibly cope with the time-varying energy capacities of mobile devices. Combining FL and SNNs is however non-trivial, particularly under wireless connections with time-varying channel conditions. Furthermore, existing multi-width SNN training algorithms are sensitive to the data distributions across devices, so are ill-suited to FL. Motivated by this, we propose a communication and energy-efficient SNN-based FL (named SlimFL) that jointly utilizes superposition coding (SC) for global model aggregation and superposition training (ST) for updating local models. By applying SC, SlimFL exchanges the superposition of multiple width configurations that are decoded as many as possible for a given communication throughput. Leveraging ST, SlimFL aligns the forward propagation of different width configurations, while avoiding the inter-width interference during backpropagation. We formally prove the convergence of SlimFL. The result reveals that SlimFL is not only communication-efficient but also can counteract non-IID data distributions and poor channel conditions, which is also corroborated by simulations.