Postmeasurement information and nonlocality of quantum state discrimination

Jinhyeok Heo, Donghoon Ha, Jeong San Kim · 2024

Read Postmeasurement information and nonlocality of quantum state discrimination by Jinhyeok Heo, Donghoon Ha, Jeong San Kim (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-17435.txt.

Introduction

Jinhyeok Heo, Donghoon Ha, and Jeong San Kim show that extra information handed over after a quantum measurement can create or destroy nonlocality, in a 2024 paper. Which of the two happens depends on how the clues are grouped. In state discrimination, distant parties guess which state was prepared; sometimes no local strategy achieves the best possible guess, and that gap is the nonlocality.

Recent work showed that revealing which subensemble the state came from can switch the gap on or off. The new result makes the effect partition-dependent: how the original ensemble is split into subensembles decides the outcome. The authors derive sufficient conditions for each direction and build explicit two-party ensembles satisfying them. The moral is that nonlocality here is not a fixed property of the states alone. It lives partly in how the question is framed after the fact.

Catalog record

CREATORJinhyeok Heo, Donghoon Ha, Jeong San Kim
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE45 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-17435

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Postmeasurement information and nonlocality of quantum state discrimination Jinhyeok Heo, Donghoon Ha, Jeong San Kim Abstract In quantum state discrimination, nonlocality arises when optimal discrimination cannot be achieved using only local operations and classical communication. It has recently been shown that postmeasurement information identifying the subensemble containing the prepared state can either annihilate or create such nonlocality. Here, we demonstrate that these effects can depend critically on how the original ensemble is partitioned into subensembles. We derive sufficient conditions under which different ensemble partitions lead to the annihilation or creation of nonlocality upon the revelation of postmeasurement information. We further construct explicit bipartite quantum state ensembles satisfying these conditions, thereby illustrating the partition-dependent nature of nonlocality in quantum state discrimination. I Introduction Quantum nonlocality is a fundamental feature of multipartite quantum systems. A paradigmatic manifestation of such nonlocality is quantum entanglement, which cannot be generated from separable states using only local operations and classical communication (LOCC) [ 1 , 2 ] . Entanglement serves as an essential resource for various tasks in multipartite quantum information processing, including quantum teleportation and quantum communication [ 3 , 4 ] . These developments have also motivated extensive efforts to characterize other forms of quantum nonlocality, such as Bell nonlocality and quantum steering [ 5 , 6 ] . Nonlocality also arises in the discrimination of multipartite quantum states. For a given ensemble of quantum states, nonlocality is said to occur when globally optimal discrimination cannot be achieved using only LOCC [ 7 ] . In general, mutually orthogonal quantum states can be perfectly discriminated by global measurements, whereas nonorthogonal states cannot [ 8 , 9 , 10 , 11 ] . Nevertheless, some sets of multipartite orthogonal quantum states cannot be perfectly discriminated using only LOCC [ 12 , 13 , 14 ] . Moreover, certain ensembles of nonorthogonal quantum states cannot be optimally discriminated by LOCC [ 15 , 16 , 17 ] . Furthermore, LOCC can activate nonlocality in state discrimination by transforming certain locally distinguishable ensembles into locally indistinguishable ones [ 18 , 19 , 20 , 21 ] . Recently, it was shown that nonlocality in quantum state discrimination can be either annihilated or created when postmeasurement information (PI) identifying the subensemble containing the prepared state is provided [ 22 , 23 , 25 , 24 ] . Some ensembles that exhibit nonlocality without PI become optimally distinguishable by LOCC once the prepared subensemble is revealed. In this case, PI annihilates the nonlocality in state discrimination. Conversely, there exist ensembles that can be optimally discriminated by LOCC without PI but cease to be optimally distinguishable by LOCC when the prepared subensemble is revealed [ 25 , 24 ] . In this case, PI creates nonlocality. These contrasting effects naturally raise the question of whether PI about the prepared subensemble can always be used to annihilate or create nonlocality in quantum state discrimination. Here, we consider bipartite quantum state discrimination and show that the annihilation or creation of nonlocality by PI can depend on how the original ensemble is partitioned into subensembles. In particular, even when PI associated with one ensemble partition annihilates or creates nonlocality, PI associated with another partition need not have the same effect. We establish sufficient conditions for this partition-dependent annihilation or creation of nonlocality (Theorems 3 and 4 ) and construct bipartite quantum state ensembles satisfying these conditions (Examples 1 and 2 ). This paper is organized as follows. In Sec. II , we review the definitions and relevant properties of separable measurements and entanglement witnesses used to analyze nonlocality in quantum state discrimination. We also introduce minimum-error discrimination (ME) and present several useful properties of nonlocality in ME. In Sec. III , we review ME with postmeasurement information and establish sufficient conditions for nonlocality in this setting (Theorems 1 and 2 ). In Sec. IV , we introduce the notions of annihilating and creating nonlocality by PI and derive sufficient conditions under which these effects depend on the choice of subensembles specified by PI (Theorems 3 and 4 ). We then construct four-state ensembles for which nonlocality is annihilated or created depending on how the ensemble is partitioned into two-state subensembles (Examples 1 and 2 ). Finally, in Sec. V , we summarize our results and discuss directions for future research. II Preliminary For a bipartite Hilbert space \mathcal{H}=\mathbb{C}^{d_{\rm A}}\otimes\mathbb{C}^{d_{\rm B}} , let \mathbb{H} be the set of all Hermitian operators acting on \mathcal{H} . We denote by \mathbb{H}_{+} the set of all positive-semidefinite operators in \mathbb{H} , that is, \mathbb{H}_{+}=\{E\in\mathbb{H}\,|\,\bra{v}E\ket{v}\geqslant 0~\forall\ket{v}\in\mathcal{H}\}. (1) A bipartite quantum state is described by a density operator \rho , that is, a positive-semidefinite operator \rho\in\mathbb{H}_{+} with unit trace \Tr\rho=1 . A measurement is represented by a positive operator-valued measure \{M_{i}\}_{i} , that is, a set of positive-semidefinite operators M_{i}\in\mathbb{H}_{+} satisfying the completeness relation \sum_{i}M_{i}=\mathbbm{1} , where \mathbbm{1} is the identity operator in \mathbb{H} . For the state \rho , the probability of obtaining the measurement outcome with respect to M_{j} is \Tr(\rho M_{j}) . Definition 1 . E\in\mathbb{H}_{+} is called separable if it can be described as E=\sum_{l}A_{l}\otimes B_{l}, (2) where A_{l} and B_{l} are positive-semidefinite operators acting on \mathbb{C}^{d_{\rm A}} and \mathbb{C}^{d_{\rm B}} of \mathcal{H} , respectively. II.1 Entanglement witness We denote the set of all separable operators in \mathbb{H}_{+} as \mathbb{SEP}=\{E\in\mathbb{H}_{+}\,|\,E:\mbox{separable}\}, (3) and its dual set as \mathbb{SEP}^{*} , that is, \mathbb{SEP}^{*}=\{E\in\mathbb{H}\,|\,\Tr(EF)\geqslant 0~\forall F\in\mathbb{SEP}\}. (4) An element in \mathbb{SEP}^{*} is also called block positive . A measurement \{M_{i}\}_{i} is called a separable measurement if M_{i}\in\mathbb{SEP} for all i , and a measurement is called a LOCC measurement if it can be realized by LOCC. Note that every LOCC measurement is a separable measurement [ 2 ] . Definition 2 . W\in\mathbb{H} is called an entanglement witness (EW) if \Tr(\sigma W)\geqslant 0 for any state \sigma in \mathbb{SEP} but \Tr(\rho W)<0 for some state \rho in \mathbb{H}_{+}\backslash\mathbb{SEP} , or equivalently W\in\mathbb{SEP}^{*}\backslash\mathbb{H}_{+}. (5) Throughout this paper, we only consider the situation of discriminating bipartite quantum states from the ensemble of the form, \mathcal{E}=\{\eta_{i},\rho_{i}\}_{i\in\Lambda},~\Lambda=\{1,2,3,4\}, (6) where the state \rho_{i} is prepared with the probability \eta_{i} . II.2 Nonlocality in bipartite quantum state discrimination Let us consider the quantum state discrimination of \mathcal{E} in Eq. ( 6 ) using a measurement \mathcal{M}=\{M_{i}\}_{i\in\Lambda} . Here, the detection of M_{i} means that the prepared state is guessed to be \rho_{i} . The ME of \mathcal{E} is to achieve the optimal success probability, p_{\rm G}(\mathcal{E})=\max_{\mathcal{M}}\sum_{i\in\Lambda}\eta_{i}\Tr(\rho_{i}M_{i}), (7) where the maximum is taken over all possible measurements [ 26 ] . When the available measurements are limited to LOCC measurements, we denote the maximum success probability by p_{\rm L}(\mathcal{E})=\max_{\mathrm{LOCC}\,\mathcal{M}}\sum_{i\in\Lambda}\eta_{i}\Tr(\rho_{i}M_{i}). (8) Similarly, we denote the maximum success probability over all possible separable measurements as p_{\rm SEP}(\mathcal{E})=\max_{\mathrm{Separable}\,\mathcal{M}}\sum_{i\in\Lambda}\eta_{i}\Tr(\rho_{i}M_{i}). (9) From the definitions, we have \max\{\eta_{1},\ldots,\eta_{n}\}\leqslant p_{\rm L}(\mathcal{E})\leqslant p_{\rm SEP}(\mathcal{E})\leqslant p_{\rm G}(\mathcal{E}). (10) In discriminating the states from the ensemble \mathcal{E} , quantum nonlocality occurs if the optimal success probability in Eq. ( 7 ) cannot be achieved only by LOCC measurements, that is, p_{\rm L}(\mathcal{E})<p_{\rm G}(\mathcal{E}). (11) From Inequality ( 10 ), we can easily see that Inequality ( 11 ) holds if p_{\rm SEP}(\mathcal{E})<p_{\rm G}(\mathcal{E}). (12) Moreover, the following proposition provides a condition for Inequality ( 12

Read the complete text