arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2409.08038quant-ph

迈向可扩展量子密钥分发:一种基于机器学习的Cascade协议方法

Towards Scalable Quantum Key Distribution: A Machine Learning-Based Cascade Protocol Approach

Hasan Abbas Al-Mohammed, Saif Al-Kuwari, Hashir Kuniyil, Ahmed Farouk

AI总结:

针对量子密钥分发(QKD)可扩展性不足的问题,提出将机器学习自编码器与Cascade纠错协议结合的方法,可超99%准确率预测QBER和密钥长度,大幅降低纠错时间,提升QKD系统效率与实用性,为高吞吐量可扩展QKD树立新基准。

摘要:

量子密钥分发(QKD)是实现安全通信的关键技术,它利用量子力学原理保障数据的可靠防护。然而,如何扩展QKD以满足高速实际应用的需求,仍是一项重大挑战。传统的密钥速率确定方法依赖复杂的数学模型,在效率和可扩展性方面往往存在不足。本文提出一种将机器学习(ML)技术与Cascade纠错协议相结合的方法,以提升QKD系统的可扩展性和效率。我们的基于ML的方法采用autoencoder(自编码器)框架预测量子比特误码率(QBER)和最终密钥长度,准确率超过99%。该方法大幅缩短了纠错时间,即使在输入规模较大(如数据速率高达156 Mbps)的情况下,计算时间也始终保持在较低水平。相比之下,传统方法的计算时间会随输入规模增大呈指数级增长,凸显了我们基于ML的解决方案在可扩展性上的优势。通过全面的仿真实验,我们证明该方法不仅加快了纠错过程,还优化了资源利用,使其在实际部署中更具成本效益和实用性。Cascade协议的融入还能通过基于实时QBER观测动态调整纠错策略,进一步增强系统安全性,为潜在的窃听行为提供可靠防护。本研究为可扩展、高吞吐量的QKD系统树立了新的基准,证明机器学习能够显著推动量子密码学领域的发展,为实现真正可扩展的量子通信持续演进奠定基础。

英文摘要:

Quantum Key Distribution (QKD) is a pivotal technology in the quest for secure communication, harnessing the power of quantum mechanics to ensure robust data protection. However, scaling QKD to meet the demands of high-speed, real-world applications remains a significant challenge. Traditional key rate determination methods, dependent on complex mathematical models, often fall short in efficiency and scalability. In this paper, we propose an approach that involves integrating machine learning (ML) techniques with the Cascade error correction protocol to enhance the scalability and efficiency of QKD systems. Our ML-based approach utilizes an autoencoder framework to predict the Quantum Bit Error Rate (QBER) and final key length with over 99\% accuracy. This method significantly reduces error correction time, maintaining a consistently low computation time even with large input sizes, such as data rates up to 156 Mbps. In contrast, traditional methods exhibit exponentially increasing computation times as input sizes grow, highlighting the superior scalability of our ML-based solution. Through comprehensive simulations, we demonstrate that our method not only accelerates the error correction process but also optimizes resource utilization, making it more cost-effective and practical for real-world deployment. The Cascade protocol's integration further enhances system security by dynamically adjusting error correction based on real-time QBER observations, providing robust protection against potential eavesdropping. Our research establishes a new benchmark for scalable, high-throughput QKD systems, proving that machine learning can significantly advance the field of quantum cryptography. This work continues the evolution towards truly scalable quantum communication.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑