Over-The-Air Clustered Wireless Federated Learning

TL;DR

This paper introduces OTA semi-decentralized clustered wireless federated learning algorithms that improve communication efficiency and convergence, achieving accuracy comparable to centralized methods while reducing reliance on a powerful server.

Contribution

It proposes novel OTA semi-decentralized clustered FL algorithms (CWFL and CWFL-Prox) that enhance communication efficiency and convergence in wireless federated learning without a central server.

Findings

CWFL achieves convergence to global minima at O(1/T).

CWFL's accuracy is comparable to centralized methods on MNIST and CIFAR10.

The proposed algorithms outperform single-client models in accuracy.

Abstract

Privacy and bandwidth constraints have led to the use of federated learning (FL) in wireless systems, where training a machine learning (ML) model is accomplished collaboratively without sharing raw data. While using bandwidth-constrained uplink wireless channels, over-the-air (OTA) FL is preferred since the clients can transmit parameter updates simultaneously to a server. A powerful server may not be available for parameter aggregation due to increased latency and server failures. In the absence of a powerful server, decentralised strategy is employed where clients communicate with their neighbors to obtain a consensus ML model while incurring huge communication cost. In this work, we propose the OTA semi-decentralised clustered wireless FL (CWFL) and CWFL-Prox algorithms, which is communication efficient as compared to the decentralised FL strategy, while the parameter updates converge to global minima as O(1/T) for each cluster. Using the MNIST and CIFAR10 datasets, we demonstrate the accuracy performance of CWFL is comparable to the central-server based COTAF and proximal constraint based methods, while beating single-client based ML model by vast margins in accuracy.