{"id":"arxiv/2603.04192","url":"https://arxiv.org/abs/2603.04192","tags":["quant-ph"],"score":0,"title":"AML-QKD: Adaptive Machine Learning Framework for Real-time Parameter Tuning in QKD","domain":"quant-ph","source":"arxiv","authors":["Noureldin Mohamed","Jawaher Kaldari","Saif Al-Kuwari"],"reviews":0,"subject":"quant-ph","summary":"Despite the robust security guarantees of Quantum Key Distribution (QKD), practical deployment is hindered by dynamic channel noise and complex parameter optimization. We propose AML-QKD, a protocol-agnostic machine learning framework designed to maximize the Secure Key Rate (SKR) and minimize the Quantum Bit Error Rate (QBER) across the BB84, E91, and COW protocols. AML-QKD integrates Temporal Convolutional Networks (TCNs) for short-horizon forecasting of channel fluctuations, using a Proximal Policy Optimization (PPO) agent for real-time parameter tuning, while strictly adhering to…","category":"research","comments":0,"disputed":false,"division":"physics","mentions":0,"discussed":0,"published":"2026-03-04","reactions":{}}