No Arabic abstract
Conventionally, unknown quantum states are characterized using quantum-state tomography based on strong or weak measurements carried out on an ensemble of identically prepared systems. By contrast, the use of protective measurements offers the possibility of determining quantum states from a series of weak, long measurements performed on a single system. Because the fidelity of a protectively measured quantum state is determined by the amount of state disturbance incurred during each protective measurement, it is crucial that the initial quantum state of the system is disturbed as little as possible. Here we show how to systematically minimize the state disturbance in the course of a protective measurement, thus enabling the maximization of the fidelity of the quantum-state measurement. Our approach is based on a careful tuning of the time dependence of the measurement interaction and is shown to be dramatically more effective in reducing the state disturbance than the previously considered strategy of weakening the measurement strength and increasing the measurement time. We describe a method for designing the measurement interaction such that the state disturbance exhibits polynomial decay to arbitrary order in the inverse measurement time $1/T$. We also show how one can achieve even faster, subexponential decay, and we find that it represents the smallest possible state disturbance in a protective measurement. In this way, our results show how to optimally measure the state of a single quantum system using protective measurements.
The uncertainty principle states that a measurement inevitably disturbs the system, while it is often supposed that a quantum system is not disturbed without state change. Korzekwa, Jennings, and Rudolph [Phys. Rev. A 89, 052108 (2014)] pointed out a conflict between those two views, and concluded that state-dependent formulations of error-disturbance relations are untenable. Here, we reconcile the conflict by showing that a quantum system is disturbed without state change, in favor of the recently obtained universally valid state-dependent error-disturbance relations.
The trade-off between the information gain and the state disturbance is derived for quantum operations on a single qubit prepared in a uniformly distributed pure state. The derivation is valid for a class of measures quantifying the state disturbance and the information gain which satisfy certain invariance conditions. This class includes in particular the Shannon entropy versus the operation fidelity. The central role in the derivation is played by efficient quantum operations, which leave the system in a pure output state for any measurement outcome. It is pointed out that the optimality of efficient quantum operations among those inducing a given operator-valued measure is related to Davies characterization of convex invariant functions on hermitian operators.
We demonstrate a general method to measure the quantum state of an angular momentum of arbitrary magnitude. The (2F+1) x (2F+1) density matrix is completely determined from a set of Stern-Gerlach measurements with (4F+1) different orientations of the quantization axis. We implement the protocol for laser cooled Cesium atoms in the 6S_{1/2}(F=4) hyperfine ground state and apply it to a variety of test states prepared by optical pumping and Larmor precession. A comparison of input and measured states shows typical reconstruction fidelities of about 0.95.
We investigate the optimal tradeoff between information gained about an unknown coherent state and the state disturbance caused by the measurement process. We propose several optical schemes that can enable this task, and we implement one of them, a scheme which relies on only linear optics and homodyne detection. Experimentally we reach near optimal performance, limited only by detection inefficiencies. In addition we show that such a scheme can be used to enhance the transmission fidelity of a class of noisy channels.
Entanglement is a fundamental feature of quantum mechanics, considered a key resource in quantum information processing. Measuring entanglement is an essential step in a wide range of applied and foundational quantum experiments. When a two-particle quantum state is not pure, standard methods to measure the entanglement require detection of both particles. We introduce a method in which detection of only one of the particles is required to characterize the entanglement of a two-particle mixed state. Our method is based on the principle of quantum interference. We use two identical sources of a two-photon mixed state and generate a set of single-photon interference patterns. The entanglement of the two-photon quantum state is characterized by the visibility of the interference patterns. Our experiment thus opens up a distinct avenue for verifying and measuring entanglement, and can allow for mixed state entanglement characterization even when one particle in the pair cannot be detected.