A Routing Framework for Quantum Entanglements with Heterogeneous Duration

A Routing Framework for Quantum Entanglements with Heterogeneous Duration
复制标题

DOI:
10.1109/qce57702.2023.00128
复制
发表时间:
2023-09
期刊:
2023 IEEE International Conference on Quantum Computing and Engineering (QCE)
影响因子:
--
通讯作者:
Yuhang Gan;Xiaoxue Zhang;Ruilin Zhou;Yi Liu;Chen Qian
Yuhang Gan;Xiaoxue Zhang;Ruilin Zhou;Yi Liu;Chen Qian
中科院分区:
其他
文献类型:
--
作者:
Yuhang Gan;Xiaoxue Zhang;Ruilin Zhou;Yi Liu;Chen Qian

文献摘要

相似文献

纠缠路由是量子网络网络层的一个基本问题,已成为一个日益热门的研究课题。针对现有纠缠路由算法的一个共同问题:它们采用同步单时隙模型(SynSts),要求所有纠缠的产生和交换都在一个时隙内完成。这种模型与未来的大规模量子网络不一致,后者将逐步部署异构设备:旧设备产生短时间的纠缠,而先进设备可以产生长时间的纠缠。强制长持续时间的纠缠在一个时隙中完成将导致次优的路由性能。本文提出了一种新的路由框架,包括同步多时隙(SynMts)路由模型和两种路由请求和链路状态管理算法。该框架以这样一种通用的方式设计,即为SynSts模型设计的所有现有算法仍然可以与SynMts模型一起运行,同时实现更高的性能。我们使用三个最近提出的路由算法进行实验,发现他们都取得了显着的吞吐量改善在新的框架。
Entanglement routing is a fundamental problem in the network layer of quantum networks, which has become an increasingly popular research topic. This paper addresses a common problem of existing entanglement routing algorithms: they assume the Synchronized Single-time-slot model (SynSts) that requires all entanglement generation and swapping to be completed in a single time slot. This model does not align with future large-scale quantum networks that will incrementally deploy heterogeneous devices: old ones generate short-duration entanglements while advanced devices can generate long-duration entanglements. Forcing long-duration entanglements to finish in one time slot will result in sub-optimal routing performance. This paper presents a new routing framework for quantum entanglements with heterogeneous duration, including a Synchronous Multi-time-slots (SynMts) routing model and two algorithms to manage routing requests and link states. The framework is designed in such a generalized way that all existing algorithms designed for the SynSts model can still run with the SynMts model while achieving higher performance. We conduct experiments using three recently-proposed routing algorithms and find they all achieve evident throughput improvement in the new framework.