摘要
量子生成对抗网络(quantum generative adversarial networks,QGAN)在图像处理、金融分析等领域应用中展现出了优越的性能.本文首次提出了一种基于量子生成对抗网络的量子拓扑码解码器,并应用于优化容错量子隐形传态系统.在本文中,首先构建并测试了QGAN算法的量子线路,搭建了拓扑码解码器训练模型.其次,针对拓扑码本征值数据集,设计了算法的输入输出,并训练得到高效率的解码模型.最后,构建了带有QGAN解码器的拓扑码优化量子隐形传态系统,相较于原始系统展现出更好的容错性能.在码距d=3及d=5的解码实验表明,本模型纠错成功率可以达到99.887%.在实验中,本QGAN解码器的保真度阈值约为P=0.1706,相较经典解码模型阈值约为P=0.1099,有了明显提升.另外,量子隐形传态系统在d=3拓扑码优化抗噪下,在非极化噪声阈值P<0.0607范围内具有明显的保真度提升;在d=5拓扑码优化抗噪下,在非极化噪声阈值P<0.0778范围内具有明显的保真度提升.本文提出的QGAN解码模型,结合了量子隐形传态方法,为量子深度学习的应用提供了新思路,并可应用于其他非均匀噪声处理领域.
Quantum generative adversarial networks(QGANs)have demonstrated superior performance when applied to image processing,financial analysis,and other fields.This paper proposes a quantum topological code decoder based on QGANs,which is applied to optimize fault-tolerant quantum teleportation systems.In this study,we first construct and test the quantum circuit of a QGAN algorithm,establishing a topological code decoder training model.Subsequently,algorithms are designed for the input and output of the topological code eigenvalue dataset,and an efficient decoder is trained.Finally,a topological-code-optimized quantum teleportation system with a QGAN decoder is constructed,exhibiting better fault-tolerance performance compared to the original system.Decoding experiments with a code distance d=3 and d=5 show that the error correction success rate of this model reaches 99.887%.The QGAN decoder demonstrates a fidelity threshold of P=0:1706,significantly higher than the classical decoder threshold,which is approximately P=0:1099.Furthermore,the quantum teleportation system,optimized for noise resistance under a topological code with d=3,shows a noticeable fidelity improvement within the non-polarized noise threshold range of P<0:0607,while under a topological code optimization with d=5,there is a significant fidelity improvement within the non-polarized noise threshold range of P<0:0778.The proposed QGAN decoding model,combined with quantum teleportation methods,provides a novel approach for the application of quantum deep learning,whose principles can be applied to other areas of non-uniform noise processing.
作者
李嘉鑫
史尚尚
尚瑞敏
李亚男
王志敏
顾永建
Jiaxin LI;Shangshang SHI;Ruimin SHANG;Yanan LI;Zhimin WANG;Yongjian GU(College of Information Science and Engineering,Ocean University of China,Qingdao 266100,China)
出处
《中国科学：信息科学》
CSCD
北大核心
2024年第6期1541-1557,共17页
Scientia Sinica(Informationis)
基金
山东省自然科学基金(批准号:ZR2021ZD19)
国家自然科学基金(批准号:12005212)资助项目。
