Multi-Armed Bandit Based Client Scheduling for Federated Learning

TL;DR

This paper introduces multi-armed bandit algorithms for client scheduling in federated learning to reduce latency, handling both ideal and non-ideal data scenarios with theoretical regret bounds and validated by simulations.

Contribution

It proposes novel UCB-based client scheduling algorithms for federated learning that adapt to data heterogeneity and client availability, with proven regret bounds and improved training efficiency.

Findings

Regret grows logarithmically in ideal scenarios.

Sub-linear regret growth in non-ideal scenarios.

Simulation results confirm algorithm efficiency.

Abstract

By exploiting the computing power and local data of distributed clients, federated learning (FL) features ubiquitous properties such as reduction of communication overhead and preserving data privacy. In each communication round of FL, the clients update local models based on their own data and upload their local updates via wireless channels. However, latency caused by hundreds to thousands of communication rounds remains a bottleneck in FL. To minimize the training latency, this work provides a multi-armed bandit-based framework for online client scheduling (CS) in FL without knowing wireless channel state information and statistical characteristics of clients. Firstly, we propose a CS algorithm based on the upper confidence bound policy (CS-UCB) for ideal scenarios where local datasets of clients are independent and identically distributed (i.i.d.) and balanced. An upper bound of the expected performance regret of the proposed CS-UCB algorithm is provided, which indicates that the regret grows logarithmically over communication rounds. Then, to address non-ideal scenarios with non-i.i.d. and unbalanced properties of local datasets and varying availability of clients, we further propose a CS algorithm based on the UCB policy and virtual queue technique (CS-UCB-Q). An upper bound is also derived, which shows that the expected performance regret of the proposed CS-UCB-Q algorithm can have a sub-linear growth over communication rounds under certain conditions. Besides, the convergence performance of FL training is also analyzed. Finally, simulation results validate the efficiency of the proposed algorithms.