We show (i) the existence of universal resource states for a certain class of linear Hamiltonians and (ii) the uselessness of highly entangled states for quantum metrology of linear Hamiltonians. We also show that random pure states are basically not useful even if we consider more general Hamiltonians. Since random pure states have high entanglement, this result strengthens the uselessness of highly entangled states for quantum metrology.
We show (i) the existence of universal resource states for a certain class of linear Hamiltonians and (ii) the uselessness of highly entangled states for quantum metrology of linear Hamiltonians. We also show that random pure states are basically not useful even if we consider more general Hamiltonians. Since random pure states have high entanglement, this result strengthens the uselessness of highly entangled states for quantum metrology.
We investigate which quantum states can serve as universal resources for approximate and stochastic measurement-based quantum computation in the sense that any quantum state can be generated from a given resource by means of single-qubit (local) operations assisted by classical communication. More precisely, we consider the approximate and stochastic generation of states, resulting, for example, from a restriction to finite measurement settings or from possible imperfections in the resources or local operations. We show that entanglement-based criteria for universality obtained in M. Van den Nest et al. [New J. Phys. 9, 204 (2007)] for the exact, deterministic case can be lifted to the much more general approximate, stochastic case. This allows us to move from the idealized situation (exact, deterministic universality) considered in previous works to the practically relevant context of nonperfect state preparation. We find that any entanglement measure fulfilling some basic requirements needs to reach its maximum value on some element of an approximate, stochastic universal family of resource states, as the resource size grows. This allows us to rule out various families of states as being approximate, stochastic universal. We prove that approximate, stochastic universality is in general a weaker requirement than deterministic, exact universality and provide resources that are efficient approximate universal, but not exact deterministic universal. We also study the robustness of universal resources for measurement-based quantum computation under realistic assumptions about the (imperfect) generation and manipulation of entangled states, giving an explicit expression for the impact that errors made in the preparation of the resource have on the possibility to use it for universal approximate and stochastic state preparation. Finally, we discuss the relation between our entanglement-based criteria and recent results regarding the uselessness of states with a high degree of geometric entanglement as universal resources [D. Gross et al., Phys. Rev. Lett. 102, 190501 (2009); M. J. Bremner et al., Phys. Rev. Lett 102, 190502 (2009)].
We consider bipartite entangled states that cannot outperform separable states in any linear interferometer. Then, we show that these states can still be more useful metrologically than separable states if several copies of the state are provided or an ancilla is added to the quantum system. We present a general method to find the local Hamiltonian for which a given quantum state performs the best compared to separable states. We obtain analytically the optimal Hamiltonian for some quantum states with a high symmetry. We show that all bipartite entangled pure states outperform separable states in metrology. Some potential applications of the results are also suggested.
Abstract We consider quantum metrology with several copies of bipartite and multipartite quantum states. We characterize the metrological usefulness by determining how much the state outperforms separable states. We identify a large class of entangled states that become maximally useful for metrology in the limit of large number of copies, even if the state is weakly entangled and not even more useful than separable states. This way we activate metrologically useful genuine multipartite entanglement. Remarkably, not only that the maximally achievable metrological usefulness is attained exponentially fast in the number of copies, but it can be achieved by the measurement of few simple correlation observables. We also make general statements about the usefulness of a single copy of pure entangled states. We surprisingly find that the multiqubit states presented in Hyllus et al (2010 Phys. Rev. A 82 012337), which are not useful, become useful if we embed the qubits locally in qutrits. We discuss the relation of our scheme to error correction, and its possible use for quantum metrology in a noisy environment.
This thesis aims to help in bridging the gap between the ideals of theoretical quantum metrology and its practical application. We take three major approaches to achieving this aim: to investigate the effect of noise on metrological schemes, to study the role of resources and to devise ways to engineer effective probe states from experimentally feasible elements. We find that noise is detrimental to measurement precision but that this may be overcome by thinking outside of the asymptotic regime. Considering resources, whilst entanglement may provide an advantage, it is only useful in limited circumstances and even in this case a very small entangled state may perform as well as a large one. This is also true in multiparameter metrology, where we find that it is possible for the simultaneous estimation of phase and noise to be advantageous, even when the parameters are not fully compatible. We discuss strategies for multiparameter estimation in both discrete and continuous variable formalisms. In the continuous variable formalism, we go beyond the consideration of entanglement as a resource and construct a generalised resource theory for Gaussian states and operations which has particular cases in entanglement, steerability and squeezing. We provide a measure for this resource as well as a no-go theorem for its distillation. Finally, we use a genetic algorithm to devise state engineering schemes to produce highly sensitive states for quantum metrology.
Quantum metrology, which addresses parameter estimation in quantum systems, has broad applications across science and technology. Conventional metrology protocols for multi-qubit states in the multi-parameter regime typically require highly complex quantum measurements, leading to substantial quantum-resource costs. In this work, we introduce a family of metrology protocols that use only few-qubit measurements, thereby significantly reducing the required resources. For arbitrary pure states, one of our protocols approaches the quantum Cramér-Rao bound up to an overhead in sample complexity that scales linearly with the number of qubits, irrespective of the number of parameters to be estimated. For typical Haar-random states, this overhead can be reduced to a constant. Our results build on recent advances in quantum state certification protocols with few-qubit measurements: we establish a universal connection between certification and metrology in which the precision of the certification protocol determines the metrological overhead. We also illustrate our approach through an example of Hamiltonian estimation from ground states.
Entanglement-assisted quantum communication employs preshared entanglement between sender and receiver as a resource. We apply the same framework to quantum metrology, introducing shared entanglement between the probe and the ancilla in the preparation stage and allowing entangling operations at the measurement stage, i.e., using some entangled ancillary system that does not interact with the system to be sampled. This is known to be useless in the noiseless case, but was recently shown to be useful in the presence of noise [R. Demkowicz-Dobrzanski and L. Maccone, Phys. Rev. Lett. 113, 250801 (2014); W. D\"ur, M. Skotiniotis, F. Fr\"owis, and B. Kraus, Phys. Rev. Lett. 112, 080801 (2014); E. M. Kessler, I. Lovchinsky, A. O. Sushkov, and M. D. Lukin, Phys. Rev. Lett. 112, 150802 (2014); R. Demkowicz-Dobrza\ifmmode \acute{n}\else \'{n}\fi{}ski and J. Kolodynski, New J. Phys. 15, 073043 (2013)]. Here we detail how and when it can be of use. For example, surprisingly it is useful when two channels are randomly alternated, for both of which ancillas do not help (depolarizing). We show that it is useful for all levels of noise for many noise models and propose a simple optical experiment to test these results.