Exploring the Accuracy of Interferometric Quantum Measurements under Conservation Laws

TL;DR

This paper investigates how conservation laws affect the accuracy of interferometric quantum measurements using microscopic meters, revealing fundamental links between measurement error, wavepacket properties, and the bounds set by Ozawa's theorem.

Contribution

It introduces a practical interferometric setup to analyze measurement errors under conservation laws, highlighting the influence of wavepacket shape and size on measurement precision.

Findings

Measurement error relates to the non-stationary observable and interaction duration.

Error ratio approaches 1 for Gaussian wavepackets with short wavepackets.

Long wavepackets lead to an error ratio approaching √2.

Abstract

A (target) quantum system is often measured through observations performed on a second (meter) system to which the target is coupled. In the presence of global conservation laws holding on the joint meter-target system, the Wigner-Araki-Yanase theorem and its generalizations predict a lower-bound on the measurement's error (Ozawa's bound). While practically negligible for macroscopic meters, it becomes relevant for microscopic ones. Here, we propose a simple interferometric setup, arguably within reach of present technology, in which a flying particle (a microscopic quantum meter) is used to measure a qubit by interacting with it in one arm of the interferometer. In this scenario, the globally conserved quantity is the total energy of particle and qubit. We show how the measurement error, $\varepsilon$, is linked to the non-stationary nature of the measured observable and the finite duration of the target-meter interaction while Ozawa's bound, $\varepsilon_{\mathrm B}$, only depends on the momentum uncertainty of the meter's wavepacket. When considering short wavepackets with respect to the evolution time of the qubit, we show that $\varepsilon/\varepsilon_{\mathrm B}$ is strictly tied to the position-momentum uncertainty of the meter's wavepacket and $\varepsilon/\varepsilon_{\mathrm B} \rightarrow 1$ only when employing Gaussian wavepackets. On the contrary, long wavepackets of any shape lead to $\varepsilon/\varepsilon_{\mathrm B} \rightarrow \sqrt{2}$. In addition to their fundamental relevance, our findings have important practical consequences for optimal resource management in quantum technologies.