arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.14145quant-ph
收藏

噪声环境下的量子幅度估计

Quantum Amplitude Estimation in the Presence of Noise

Eric G. Brown, Oktay Goktas, W. K. Tham

摘要:

量子振幅估计(QAE)--一种可以用比标准采样少二次方的查询来估计给定量子态的振幅的技术--是包括Grover搜索和量子Monte-Carlo方法在内的几种重要量子算法中的关键子例程。在近期有噪声的中间尺度量子(NISQ)设备中实现QAE的一个障碍是需要执行量子相位估计(QPE)-一个昂贵的过程-作为子例程。这种障碍被各种无QPE的QAE方法所消除,其中不同深度/功率的Grover查询(通常根据“时间表”)紧随其后的是测量和经典的后处理技术,如极大似然估计(MLE)。现有的分析,在这些QPE免费QAE计划的各种查询时间表的最优性,迄今为止假设无噪声系统。在这项工作中,我们分析QPE免费QAE下常见的噪声模型,可能会影响NISQ设备和各种查询计划的最优性报告在嘈杂的制度。我们证明,一个准确的噪声特性的系统,必须选择一个时间表,平衡更高的深度电路实现的更理想的性能之间的权衡,并相应地更大的积累噪声引起的错误。

英文摘要:

Quantum Amplitude Estimation (QAE) -- a technique by which the amplitude of a given quantum state can be estimated with quadratically fewer queries than by standard sampling -- is a key sub-routine in several important quantum algorithms, including Grover search and Quantum Monte-Carlo methods. An obstacle to implementing QAE in near-term noisy intermediate-scale quantum (NISQ) devices has been the need to perform Quantum Phase Estimation (QPE) -- a costly procedure -- as a sub-routine. This impediment was lifted with various QPE-free methods of QAE, wherein Grover queries of varying depths / powers (often according to a "schedule") are followed immediately by measurements and classical post-processing techniques like maximum likelihood estimation (MLE). Existing analyses as to the optimality of various query schedules in these QPE-free QAE schemes have hitherto assumed noise-free systems. In this work, we analyse QPE-free QAE under common noise models that may afflict NISQ devices and report on the optimality of various query schedules in the noisy regime. We demonstrate that, given an accurate noise characterization of one's system, one must choose a schedule that balances the trade-off between the greater ideal performance achieved by higher-depth circuits, and the correspondingly greater accumulation of noise-induced error.

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。

↑