A quantum network distributes quantum entanglements between remote nodes, which is key to many quantum applications. However, unavoidable noise in quantum operations could lead to both low throughput and low quality of entanglement distribution. This paper aims to address the simultaneous exponential degradation in throughput and quality in a buffered multi-hop quantum network. Based on an end-to-end fidelity model with worst-case (isotropic) noise, we formulate the high-fidelity remote entanglement distribution problem for a single source-destination pair, and prove its NP-hardness. To address the problem, we develop a fully polynomial-time approximation scheme for the control plane of the quantum network, and a distributed data plane protocol that achieves the desired long-term throughput and worst-case fidelity based on control plane outputs. To evaluate our algorithm and protocol, we develop a discrete-time quantum network simulator. Simulation results show the superior performance of our approach compared to existing fidelity-agnostic and fidelity-aware solutions.

翻译：[translated abstract in Chinese] 量子网络可在远程节点间分发量子纠缠，这是众多量子应用的关键。然而，量子操作中不可避免的噪声会导致纠缠分发的吞吐量与保真度均处于较低水平。本文旨在解决缓冲型多跳量子网络中吞吐量与保真度同时呈指数级恶化的问题。基于最坏情况（各向同性噪声）下的端到端保真度模型，我们针对单源-单目标对提出了高保真远程纠缠分发问题，并证明其NP难解性。为解决该问题，我们为量子网络的控制平面开发了完全多项式时间近似方案，并设计了分布式数据平面协议，该协议基于控制平面输出可实现期望的长期吞吐量与最坏情况保真度。为评估算法与协议，我们开发了离散时间量子网络仿真器。仿真结果表明，相较于现有保真度无关与保真度感知方案，本方法具有优越性能。