A2G-QFL: Adaptive Aggregation with Two Gains in Quantum Federated learning

2025-12-03

AI Summary

Quantum federated learning (QFL) deployed across heterogeneous classical networks and quantum hardware faces critical challenges—including non-uniform client quality, stochastic fluctuations in quantum teleportation fidelity, and manifold mismatch between local and global models—leading to unstable convergence and degraded accuracy. To address these issues, we propose A2G-QFL, an adaptive aggregation framework featuring a novel dual-gain mechanism: geometric gain aligns parameter manifolds via Riemannian geometry, while QoS gain dynamically weights client contributions based on real-time fidelity, latency, and device stability. Built upon a quantum-classical hybrid testbed, A2G-QFL integrates these metrics into a convergence-guaranteed adaptive update rule. Experimental results demonstrate that A2G-QFL significantly improves training stability and final model accuracy under realistic noise and heterogeneity, outperforming both FedAvg and state-of-the-art QoS-aware aggregation methods.

Technology Category

• Machine Learning: Quantum Machine Learning
• Multiagent Systems: Adversarial Agents
• Search and Optimization: Learning to Search

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Search and Retrieval-Augmented AI: Agentic search
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work

Abstract

Federated learning (FL) deployed over quantum enabled and heterogeneous classical networks faces significant performance degradation due to uneven client quality, stochastic teleportation fidelity, device instability, and geometric mismatch between local and global models. Classical aggregation rules assume euclidean topology and uniform communication reliability, limiting their suitability for emerging quantum federated systems. This paper introduces A2G (Adaptive Aggregation with Two Gains), a dual gain framework that jointly regulates geometric blending through a geometry gain and modulates client importance using a QoS gain derived from teleportation fidelity, latency, and instability. We develop the A2G update rule, establish convergence guarantees under smoothness and bounded variance assumptions, and show that A2G recovers FedAvg, QoS aware averaging, and manifold based aggregation as special cases. Experiments on a quantum classical hybrid testbed demonstrate improved stability and higher accuracy under heterogeneous and noisy conditions.

Problem

• Addresses performance degradation in quantum federated learning
• Adapts aggregation to geometric mismatch and client heterogeneity
• Enhances stability and accuracy in noisy quantum-classical networks

Innovation

• Dual gain framework for adaptive aggregation
• Geometry and QoS gains regulate blending and importance
• Convergence guarantees under smoothness and variance assumptions