Cumulative Utility Parity for Fair Federated Learning under Intermittent Client Participation

TL;DR

This paper introduces a fairness principle for federated learning that ensures clients receive comparable long-term benefits, addressing issues caused by intermittent participation and improving long-term representation parity.

Contribution

It proposes cumulative utility parity and availability-normalized utility to promote fair long-term client benefit in federated learning with intermittent participation.

Findings

Significantly improves long-term fairness among clients.

Maintains high overall model performance.

Effectively handles temporally skewed, non-IID data.

Abstract

In real-world federated learning (FL) systems, client participation is intermittent, heterogeneous, and often correlated with data characteristics or resource constraints. Existing fairness approaches in FL primarily focus on equalizing loss or accuracy conditional on participation, implicitly assuming that clients have comparable opportunities to contribute over time. However, when participation itself is uneven, these objectives can lead to systematic under-representation of intermittently available clients, even if per-round performance appears fair. We propose cumulative utility parity, a fairness principle that evaluates whether clients receive comparable long-term benefit per participation opportunity, rather than per training round. To operationalize this notion, we introduce availability-normalized cumulative utility, which disentangles unavoidable physical constraints from avoidable algorithmic bias arising from scheduling and aggregation. Experiments on temporally skewed, non-IID federated benchmarks demonstrate that our approach substantially improves long-term representation parity, while maintaining near-perfect performance.