Q-ANCHOR: Federated Quantum Learning with ZNE-guided Correction

2026-05-28

AI Summary

This work addresses the dual drift problem in quantum federated learning, arising from client drift due to non-IID data and gradient bias induced by noisy quantum hardware. To tackle this challenge, the authors propose Q-ANCHOR, a novel aggregation framework that integrates a server-side anchoring mechanism guided by zero-noise extrapolation (ZNE) with a stateful client correction strategy. The study is the first to theoretically reveal that quantum hardware bias introduces an irreducible error floor under standard federated averaging, and accordingly presents the first quantum-aware aggregation algorithm capable of simultaneously mitigating both classical client drift and quantum-induced bias. Theoretical analysis and empirical evaluations demonstrate that Q-ANCHOR significantly enhances training stability, effectively lowers the error floor, and improves model convergence.

Abstract

Quantum Federated Learning (QFL) offers a promising framework to train quantum models across distributed clients while keeping data strictly local. Due to its simplicity and low communication overhead, Federated Averaging (FedAvg) is the standard aggregation choice in QFL literature. However, deploying QFL on practical hardware exposes a severe double-drift phenomenon: the global model is simultaneously derailed by client drift from non-IID data and hardware bias from noisy quantum gradient estimates. In this work, we first analyze the convergence of FedAvg under these realistic conditions, mathematically demonstrating that quantum hardware bias creates a persistent error floor that standard averaging cannot correct. To overcome this limitation, we propose Q-ANCHOR, a quantum-aware federated aggregation architecture that anchors server updates with zero-noise extrapolation while applying stateful client correction to suppress both client drift and hardware-induced bias. Our convergence theory proves that Q-ANCHOR successfully mitigates classical client drift while actively reducing the hardware-bias floor. Experimental results demonstrate that Q-ANCHOR achieves significantly more stable training than conventional FL baselines.

Problem

• Quantum Federated Learning
• client drift
• hardware bias
• double-drift phenomenon
• noisy quantum gradient

Innovation

• Quantum Federated Learning
• Zero-Noise Extrapolation
• Client Drift
• Hardware Bias
• Federated Aggregation