POSTER: Coherence Time-Aware Predictive Filtering for Entanglement Verification in Hybrid Quantum-Classical Networks

Abstract

In hybrid quantum-classical networks, entanglement becomes usable only if classical verification messages arrive in time; yet they share links with ordinary traffic, so queuing delay may render the associated entanglement unusable. We propose QPFD, a coherence time-aware predictive filtering framework that estimates arrival-time usability and removes verification messages unlikely to preserve usable entanglement. QPFD converts these constraints into packet-level deadlines and applies predictive filtering at both ingress and egress stages of a programmable switch pipeline. Simulation results show that QPFD achieves the highest usable entanglement ratio across varying generation rates and fidelity thresholds.