论文标题
使用随机步行进行迭代阶段估计
Using Random Walks for Iterative Phase Estimation
论文作者
论文摘要
近年来，用于量子相估计的算法已经有很大的发展。在这项工作中，我们为在线贝叶斯阶段估算提供了一种新的方法，该方法可以实现Heisenberg Limited缩放量表，该缩放比例比现有的贝叶斯方法所需的误差耐受性呈指数级的经典处理时间。 这实际上意味着我们可以在CPU上进行微秒的更新，而不是现有粒子滤波器方法的毫秒。我们的方法假设先前的分布是高斯，并利用事实，当选择最佳实验时，先前分布的平均值由随机步行者的位置给出，其移动的位置由测量结果决定。然后，我们从基于Fisher信息的参数中争论我们的算法对数据提供了近乎最佳的分析。这项工作表明，在线贝叶斯推论是实用，高效的，并且可以在现代FPGA驱动的自适应实验中部署。
In recent years there has been substantial development in algorithms for quantum phase estimation. In this work we provide a new approach to online Bayesian phase estimation that achieves Heisenberg limited scaling that requires exponentially less classical processing time with the desired error tolerance than existing Bayesian methods. This practically means that we can perform an update in microseconds on a CPU as opposed to milliseconds for existing particle filter methods. Our approach assumes that the prior distribution is Gaussian and exploits the fact, when optimal experiments are chosen, the mean of the prior distribution is given by the position of a random walker whose moves are dictated by the measurement outcomes. We then argue from arguments based on the Fisher information that our algorithm provides a near-optimal analysis of the data. This work shows that online Bayesian inference is practical, efficient and ready for deployment in modern FPGA driven adaptive experiments.
