面向多后端的量子资源虚拟化调度框架OA

To address the problems such as low resource utilization and insufficient circuit execution flexibility caused by the current static resource allocation mode of quantum cloud services,a multi-backend quantum resource virtualization scheduling framework is proposed.Firstly,a full-process scheduling system covering backend resource integration,virtual resource management,allocation,and mapping execution is constructed in the framework,which enables unified scheduling of multiple quan‑tum backend resources.Secondly,a quantum chip virtual partitioning mechanism based on measure‑ment and control readout lines is designed to achieve efficient abstraction of physical quantum re‑sources into virtual resources and establish a global unified virtual resource pool.Finally,a virtual re‑source allocation algorithm is proposed,which comprehensively considers key attributes such as re‑source status and gate operation fidelity,supports dynamic merging of adjacent virtual resources,and can adapt to the execution requirements of circuits of different scales.Trials results show that under high-load scenarios,the scheme can increase quantum cloud throughput by more than 3 times,and the optimization effect of user waiting time is better than that of existing multi-programming design schemes,which effectively improves quantum resource utilization and circuit execution flexibility.