Federated Learning by Utility-Constrained Stochastic Aggregation for Improving Rational Participation

2026-05-18

AI Summary

This work addresses the challenge in federated learning where rational clients may withdraw from collaboration due to insufficient local utility, leading to degraded global model performance or even training collapse. To mitigate this issue, we propose FedUCA, a stochastic aggregation framework under utility constraints. FedUCA is the first to explicitly incorporate client participation feasibility into the aggregation mechanism by introducing utility-aware constraints at the server, which dynamically adjusts the weighting of client updates to ensure their utilities meet a minimum threshold. This approach jointly optimizes global model convergence and local client incentives. Experimental results demonstrate that FedUCA significantly improves client retention rates and achieves superior global model performance compared to existing baselines on standard benchmark datasets.

Abstract

Federated Learning (FL) algorithms implicitly assume that clients passively comply with server-side orchestration by sharing local model updates upon server request. However, this overlooks an important aspect in real-world cross-silo environments: clients are often rational agents who may prioritize their utilities such as local model performance over that of the global model. In settings with significant statistical heterogeneity, rational clients may opt out of the federation if the perceived benefits of collaboration fail to meet their local utility thresholds. Such attrition degrades the global model performance and can lead to the collapse of the federated training process. In this work, we introduce FedUCA, (Federated Learning by Utility-Constrained Stochastic Aggregation for Improving Rational Participation), a framework that formalizes the server's role as an optimizer seeking to maximize global model performance by sustaining client participation. We substantiate our framework through extensive experiments on standard datasets demonstrating that by prioritizing participation feasibility, FedUCA achieves significantly higher client retention and, consequently, a superior global model performance.

Problem

• Federated Learning
• Rational Participation
• Client Attrition
• Statistical Heterogeneity
• Utility Constraints

Innovation

• Federated Learning
• Rational Participation
• Utility-Constrained Aggregation
• Client Retention
• Statistical Heterogeneity