Where Can Buy Temporal Quantum Correlations and Hidden Variable Models 1st Edition Costantino Budroni (Auth.) Ebook With Cheap Price
Where Can Buy Temporal Quantum Correlations and Hidden Variable Models 1st Edition Costantino Budroni (Auth.) Ebook With Cheap Price
Where Can Buy Temporal Quantum Correlations and Hidden Variable Models 1st Edition Costantino Budroni (Auth.) Ebook With Cheap Price
com
https://textbookfull.com/product/temporal-quantum-
correlations-and-hidden-variable-models-1st-
edition-costantino-budroni-auth/
https://textbookfull.com/product/hidden-semi-markov-models-theory-
algorithms-and-applications-1st-edition-yu/
textbookfull.com
https://textbookfull.com/product/automotive-fdi-in-emerging-europe-
shifting-locales-in-the-motor-vehicle-industry-1st-edition-a-j-jacobs-
auth/
textbookfull.com
https://textbookfull.com/product/halstead-billionaire-brothers-the-
complete-series-boxset-1st-edition-lauren-wood/
textbookfull.com
https://textbookfull.com/product/women-s-studies-the-basics-2nd-
edition-bonnie-g-smith/
textbookfull.com
https://textbookfull.com/product/climate-clothing-and-agriculture-in-
prehistory-linking-evidence-causes-and-effects-ian-gilligan/
textbookfull.com
The Dark Side of the Workplace Managing Incivility 1st
Edition Annette B. Roter
https://textbookfull.com/product/the-dark-side-of-the-workplace-
managing-incivility-1st-edition-annette-b-roter/
textbookfull.com
Springer Theses
Recognizing Outstanding Ph.D. Research
Costantino Budroni
Temporal Quantum
Correlations and
Hidden Variable
Models
Springer Theses
The series “Springer Theses” brings together a selection of the very best Ph.D.
theses from around the world and across the physical sciences. Nominated and
endorsed by two recognized specialists, each published volume has been selected
for its scientific excellence and the high impact of its contents for the pertinent field
of research. For greater accessibility to non-specialists, the published versions
include an extended introduction, as well as a foreword by the student’s supervisor
explaining the special relevance of the work for the field. As a whole, the series will
provide a valuable resource both for newcomers to the research fields described,
and for other scientists seeking detailed background information on special
questions. Finally, it provides an accredited documentation of the valuable
contributions made by today’s younger generation of scientists.
Temporal Quantum
Correlations and Hidden
Variable Models
Doctoral Thesis accepted by
University of Siegen, Germany
123
Author Supervisor
Dr. Costantino Budroni Prof. Otfried Gühne
University of Siegen University of Siegen
Siegen Siegen
Germany Germany
v
vi Abstract
Finally, we discuss a possible application of the above results for the con-
struction of dimension witnesses, i.e., as a certification of the minimal dimension
of the Hilbert spaces needed to explain the arising of certain quantum correlations.
Supervisor’s Foreword
vii
viii Supervisor’s Foreword
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Hidden Variable Theories. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2 Local Hidden Variables and Bell’s Theorem. . . . . . . . . . . . . . . . 6
1.2.1 Local Hidden Variables . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.2.2 CHSH Inequality and Bell’s Theorem . . . . . . . . . . . . . . . 7
1.2.3 Experimental Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.3 Noncontextual Hidden Variables
and Kochen-Specker Theorem . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.3.1 Noncontextual Hidden Variable Theories . . . . . . . . . . . . . 10
1.3.2 Kochen and Specker’s Original Problem . . . . . . . . . . . . . 12
1.3.3 Kochen-Specker Theorem . . . . . . . . . . . . . . . . . . . . . . . 14
1.3.4 State-Independent Contextuality . . . . . . . . . . . . . . . . . . . 15
1.4 Macrorealist Theories and Leggett-Garg Inequalities . . . . . . . . . . 17
1.4.1 Macrorealist Theories . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.4.2 Continuous Variables. . . . . . . . . . . . . . . . . . . . . . . . . . . 19
1.4.3 Quantum Violations. . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
1.5 Correlations Polytopes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.5.1 Definition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
1.5.2 Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
1.6 Tsirelson Bound . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
1.6.1 Original Argument . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
1.6.2 Semidefinite Programming Approach . . . . . . . . . . . . . . . . 26
1.7 Linear and Semidefinite Programming . . . . . . . . . . . . . . . . . . . . 29
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2 Noncontextuality Inequalities from Variable Elimination . . . . . . . . 35
2.1 Extension of Measures and Consistency Conditions . . . . . . . . . 36
2.2 Bell Inequalities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
2.2.1 CHSH Polytope from Bell-Wigner Polytope . . . . . . . . . . 37
2.2.2 Bipartite (2, n) Scenario . . . . . . . . . . . . . . . . . . . . . . . . 39
2.2.3 Two Parties, Three Settings . . . . . . . . . . . . . . . . . . . . . 39
xi
xii Contents
The notion of probability is intimately related to the not ion of uncertainty, and the
latter arises in the description of physical systems at various levels and in different
ways. In particular, the probabilistic structure arising in quantum mechanics (QM)
has been recognized to be of a rather different kind with respect to its classical
counterpart.
The formalism of classical mechanics is based on physical quantities which are
assumed to have a clear empirical meaning (e.g., position and velocity of a point
particle) and uncertainty only arises as a consequence of practical limitations (e.g.,
finite precision of measurement apparatuses) and can also be included in a systematic
way in the description of a physical system, as it happens in statistical mechanics.
More precisely, this is done by means of a probability measure, namely, an average
over states with precisely determined values for all the physical quantities, describing
the relative frequencies for appearance of such values on systems that have been
subjected to the same preparation.
The situation is much more complex in QM. In fact, a problem of interpreta-
tion arises as a consequence of the lack of an apparent and unambiguous empirical
meaning of the elements of formalism, e.g., self-adjoint operators and state vectors.
An interpretation provides a set of rules allowing for a derivation of experimental
predictions from the formalism, e.g., rules associating self-adjoint operators with
experimental apparatuses, state vectors with preparation procedures, eigenvalues
with outcomes of experiments. In particular, the predictions are restricted to well-
defined experimental situations, avoiding some of the idealizations implicit in clas-
sical mechanics. An illustrative example of this attitude is given by Peres’ claim that
“quantum phenomena do not occur in a Hilbert space, they occur in a laboratory” [1].
Such a separation between formalism and interpretation is ultimately due to dif-
ficulties of a realistic description of quantum phenomena, i.e., a description in terms
of an “actual state of affairs”. Whether such an interpretative caution is justified, i.e.,
whether QM can be formulated in terms of classical probability theory, is still an
open question, known as the hidden variable problem.
question to identify and quantify the resources needed for quantum information
processing tasks. For instance, in quantum communication tasks, the impossibility
of a local hidden variable description, also named nonlocality, has been proven to be
a fundamental resource for secure quantum key distribution among distant parties
[12]. From the point of view of quantum computation, e.g., in the measurement-
based quantum computation model, such a locality restriction is unnecessary. In this
framework, contextuality, i.e., the impossibility of a noncontextual hidden variable
model, has been proven to be a fundamental resource for computation [13, 14].
In this thesis, we consider the problem of characterizing the ranges of values for
probabilities in different kinds of hidden variable theories as well as in QM, with
particular emphasis on the temporal scenario, and discuss possible applications of
the results (e.g., dimension witnesses).
More in details, the thesis is structured as follows. In the present chapter, we first
recall the basic definitions and properties of the three main hidden variable theories
mentioned above, namely, local, noncontextual, and macroscopic realist theories,
and how their corresponding ranges for probabilities can be computed in the unified
framework of Pitowsky’s correlation polytope [7, 15]. Such a method provides a set
of necessary and sufficient conditions for the existence of a hidden variable model,
expressed in terms of a system of linear inequalities, as a solution of a geometrical
problem known as the hull problem. Similarly, we discuss an approach for computing
quantum bounds for the Bell and contextuality scenarios.
Despite the full generality of Pitowsky’s method, and the existence of algorithms
for solution of the hull problem, computing such a minimal set of conditions is a non
trivial task. In fact, the complexity of the polytope grows rapidly with the number of
measurement settings and outcomes (e.g., the number of vertices is exponential in
the number of settings), and a direct computation has been performed only in simple
cases.
To overcome this problem, in Chap. 2 we develop an alternative method based on
the analysis of probability models for subsets of variables that are subsequently com-
bined imposing some consistency conditions on their intersection. We then proceed
to show the advantages of our method, both for analytical and numerical computation,
in some non-trivial scenarios.
In Chap. 3, we analyze the case of contextuality scenarios where each state gives
rise to the same violation of a given noncontextuality inequality, also known as state-
independent contextuality (SIC) scenarios. Due to the high number of measurement
settings involved in such scenario, the correlation polytope approach is usually inap-
plicable. We then define optimal inequalities, for a given SIC scenario, in terms of
the maximal ratio between the quantum and the classical value, and show that such
an optimization can be solved via linear programming, and thus efficiently with stan-
dard numerical techniques and with the optimality of the solution guaranteed. We
discuss the most fundamental SIC scenarios and we found that the corresponding
optimal inequalities are also facets of the associated correlation polytope.
In Chap. 4, we discuss the computation of quantum bounds for temporal correla-
tions, namely, for sequences of quantum projective measurements. We provide a gen-
eral method for computing such bounds that is based on semidefinite programming.
4 1 Introduction
unconstrained classical logical structure can be assumed, thus reducing the problem
only to the probabilistic description. More details can be found in Sect. 1.3.
In the analysis of the quantum versus classical predictions that follows, we shall
consider only the case of projective measurement. The reason is that, while the
generalization to positive-operator valued measure (POVM) can be easily done for
locality scenarios, for the other two cases it is certainly problematic. In fact, for
noncontextuality scenarios the correct notion of compatible measurements for POVM
is still under debate [23] and for macrorealist theories the use of POVM explicitly
contradicts the noninvasive measurability assumption [24–26].
Local hidden variable theories are classical theories that attempt to describe the
statistics of measurements performed on distant systems. The locality condition,
therefore, amounts to a statistical independence for the probabilities for outcomes
on separated systems once conditioned on the hidden variable.
To better introduce the main ideas involved, let us discuss the simplest measurement
scenario. Consider two experimenters, Alice and Bob, performing measurements
on two distant system. Alice can chose between two measurements, let us denote
them as x ∈ {0, 1}, with outcome a ∈ {−1, 1}, and similarly for Bob, i.e., two
measurements y ∈ {0, 1} with outcome b ∈ {−1, 1}. A schematic representation of
the measurement scenario is given in Fig. 1.1.
A LHV model is defined as probability distribution for the joint probabilities
P(ab|x y), i.e., the probability of getting the outcomes a and b given that Alice
measure x and Bob measure y, of the form
P(ab|x y) = P(λ)P(a|x, λ)P(b|y, λ) dλ. (1.3)
Fig. 1.1 Schematic representation of a Bell measurement scenario. A source S produces two
entangled particles that travel to two experiment sites, A and B. The two experimenters, Alice and
Bob, can choose their measurement settings, respectively, x and y, and get an outcome, respectively,
a and b
1.2 Local Hidden Variables and Bell’s Theorem 7
Once the hidden variable is known, the joint probability for outcomes a and b
factorizes, implying that the two variables are independent once conditioned on λ.
Without loss of generality, since λ can be chosen arbitrarily, all the indeterminacy
left in the variables a and b after conditioning on λ can be removed by redefining
the variable λ to include it. As a consequence, P(a|x, λ) and P(b|y, λ) can be seen
as deterministic functions of λ.
Notice also that in Eq. (1.3) it is implicitly assumed that the probability distribution
for the hidden variable λ does not depend on the choice of the measurement settings
x, y, an assumption is usually called free will. The origin of the name can be easily
understood by noticing that, by the definition of conditional probability,
P(λ|x, y) = P(λ) for all λ, x, y ⇐⇒ P(x, y|λ) = P(x, y) for all λ, x, y. (1.4)
Equation (1.4) implies that the experimenter is free to choose the to measure x
and y, i.e., her choice is not “influenced” by the hidden variable λ. The free will
assumption implies that λ, and consequently the a and b, must be interpreted as
statistical properties of the system that are (partially) revealed by the measurement
apparatus.
To summarize, the assumptions defining a LHV theory are the following
R Realism: Observables represent well defined properties of the system, which
are just revealed by the measurement process. In the probabilistic description
of Eq. (1.3), they are fixed once the hidden variable λ is fixed.
Loc Locality: There is a maximum speed at which information propagates. Events
in space-like separated regions cannot be in a relation of causal influence. In the
probabilistic description of Eq. (1.3), probabilities for measurements on distant
systems are statistically independent once conditioned on the hidden variable λ.
FW Free will: The experimenter is able to choose the measurement settings “freely”,
or, in simpler terms, the source of randomness used for the choice of the mea-
surement settings is independent of the source of randomness of the system
preparation. In the probabilistic description of Eq. (1.3), the probability distrib-
ution of the hidden variable λ is independent of the choice of the measurement
settings.
The form (1.3) for the probability distribution allows us to compute bounds, usually
expressed as linear inequalities, on the correlations among different outcomes. The
most celebrated is the Clauser-Horne-Shimony-Holt (CHSH) inequality [27]. Let us
denote by A0 , A1 the {−1, 1}-valued measurement settings for Alice, and by B0 , B1
the {−1, 1}-valued measurement settings for Bob, the CHSH inequality reads
where Ai B j denotes the correlation between Ai and B j , i.e., the expectation value
of the product of their outcomes.
Such a bound can be easily proven as follows. Let us define for Alice’s measure-
ments f A x (λ) = P(+1|x, λ) − P(−1|x, λ), and similarly f B y for Bob’s measure-
ment. The functions f A x , f B y are, therefore, deterministic functions of λ that fix the
measurement outcomes ±1 (note that P(+1|x, λ) + P(−1|x, λ) = 1, and similarly
for Bob). Equation (1.5), can therefore be rewritten using Eq. (1.3) as
A0 B0 + A0 B1 + A1 B0 − A1 B1 = P(λ) f A0 (λ) f B0 (λ) dλ + P(λ) f A0 (λ) f B1 (λ)dλ
+ P(λ) f A1 (λ) f B0 (λ) dλ − P(λ) f A1 (λ) f B1 (λ)dλ
=
P(λ) f A0 (λ) f B0 (λ) + f A0 (λ) f B1 (λ) + f A1 (λ) f B0 (λ) − f A1 (λ) f B1 (λ) dλ
= P(λ) f A0 (λ)( f B0 (λ) + f B1 (λ)) + f A1 (λ)( f B0 (λ) − f B1 (λ)) dλ
≤ P(λ) max f A0 (λ)( f B0 (λ) + f B1 (λ)) + f A1 (λ)( f B0 (λ) − f B1 (λ)) dλ = 2 P(λ) dλ = 2
λ
(1.6)
Proof Since Eq. (1.5) has been derived from the assumption of a local hidden variable
theory, it is sufficient to provide some quantum mechanical correlations violating the
bound.
Consider two spin-1/2 particles in the singlet state
1
|ψ = √ (|01 − |10), (1.7)
2
(1)
where |0, |1 denote the eigenstate of σz . By defining Alice’s observables A0 = σz
and A1 = σx(1) , where the superscript (1) denotes the action of the operator on the first
(2) (2) (2) (2)
particle, and Bob’s observables B0 = √1 (σz + σx ) and B1 = √1 (σz − σx ),
2 2
we obtain
4 √
A0 B0 + A0 B1 + A1 B0 − A1 B1 = √ = 2 2 > 2 (1.8)
2
1.2 Local Hidden Variables and Bell’s Theorem 9
The experimental progress in quantum optics during the 1960s, in particular the pos-
sibility of created pairs of photons entangled in polarization using atomic cascades,
allowed for the first tests of Bell inequalities. In 1972, three year after Clauser-Horne-
Shimony-Holt original proposal [27], Freedman and Clauser performed the first test
and reported a violation of the CHSH inequality by six standard deviations [28].
Freedman and Clauser’s experiment was followed by others [29–31] that share
all the same problem: The experiments were performed with a static setup in which
the polarized were held fixed. In this scenario, one can design a local hidden variable
model where the detector on one side is “aware” of the measurement setting chosen on
the other side (see the discussion on the free will assumption above). This possibility,
preventing a definite answer to the LHV problem, has been named locality loophole.
To overcome this problem, Aspect et al. [32] introduced time-varying polarization
analyzers in the experiment. With this setup, the settings were changed during the
flight of the particle in such a way that the change of orientation on one side and
the detection event on the other side were separated by a spacelike interval. This,
together with the use of independent source of randomness for the change of the
settings, justifies the free will assumption and close the locality loophole.
All the above experiments, however, were subjected to the detection loophole,
namely, the possibility of a local hidden variable model explaining the observed
correlations in terms of the statistics of the undetected events. More precisely, given
the low efficiency of photon detectors (typically around 10–20 %), one can refute
local hidden variable theories only by assuming that the fraction of detected events is
a valid representative of the whole sample (the so-called fair sampling assumption),
or, equivalently, that the probability of detecting is independent of choice of the
measurement settings.
The detection loophole in Bell experiments has been first closed by Rowe et al.
with entangled trapped ions [33], however, such an experiment was still subjected
to the locality loophole. Recently, by using highly efficient photon detectors, the
detection loophole has been closed in a photon experiment [34], thus showing that
photons can, in principle, allow for a loophole-free Bell test, albeit such a test has
not been performed yet.
In the case in which one assumes no constraint on the logical structure of the HV
theory, a definition of NCHV theory similar to the one presented above for LHV
theory can be given as follows
1.3 Noncontextual Hidden Variables and Kochen-Specker Theorem 11
where Ai are measurements with outcomes −1 and 1, and the measurements in the
same mean value are compatible, i.e., are represented in quantum mechanics by
commuting operators. The classical bound 3 can be proven, in analogy with the
CHSH case above, just by trying all possible ±1 noncontextual assignments to the
observables Ai .
As opposed to Bell inequalities, there is no bipartition of the set of observables
such that every observable in one part is compatible with every observable of the
other. Consequently, Eq. (1.10) cannot be interpreted as a Bell inequality: The mea-
surements must be performed on a single system. √
On a three-level system Eq. (1.10) can be violated up to 5 − 4 5 ≈ −3.94 with
state |ψ = (1, 0, 0) and measurement settings A j = 2|v j v j | − 1 as depicted in
Fig. 1.2, namely, |v j = (cos θ, sin θ cos[ jπ4/5], sin θ sin[ jπ4/5]) with cos2 θ =
cos(π/5)/(1 + cos(π/5)).
Kochen and Specker’s original approach [5] focused on a more strict notion of NCHV.
More precisely, it focused on reproducing also the state-independent predictions
of QM, namely, those given by functional relations between commuting quantum
observables.
As opposed to Birkhoff and von Neumann’s approach to quantum logic [36], but
rather following the same approach as Gleason [22], Kochen and Specker discussed
the possibility of reproducing just the logical relations between compatible measure-
ments, since such relations can be tested in joint measurement scenario and have a
clear experimental meaning. A discussion of this point can be found in [4].
In mathematical terms, the above notion of NCHV is captured by Kochen and
Specker’s definition of partial Boolean algebra [5], and its subsequent extension to
include probabilistic predictions [37]. Without loss of generality, we can consider
only prediction for projectors, since the outcome probabilities for any observable
can be recovered from those of its spectral decomposition. We refer to the definitions
given in [37].
A partial Boolean algebra (PBA) is a set Xtogether with a non-empty family F
of Boolean algebras, F ≡ {Bi }i∈I , such that i Bi = X , that satisfy
(P1 ) for every Bi , B j ∈ F, Bi ∩ B j ∈ F and the Boolean operations (∩i , ∪i ,ci ),
(∩ j , ∪ j ,c j ) of Bi and B j coincide on it.
Without loss of generality we can also assume the property
(P2 ) for all Bi ∈ F, each Boolean subalgebra of Bi belongs to F.
By (P1 ), Boolean operations, when defined, are unique and will be denoted by
(∩, ∪,c ); we shall denote a partial Boolean algebra by (X, {Bi }i∈I ), or simply by
{Bi }i∈I . In the following we shall consider only finite partial Boolean algebras.
Given a partial Boolean algebra (X, {Bi }), a state is defined as a map f : X −→
[0, 1], such that f |B i is a normalized measure on the Boolean algebra Bi for all i.
Equivalently, a state is given by a collection of compatible probability measures {μi },
i.e., measures coinciding on intersections of Boolean algebras, one for each Bi .
A partial probability theory (PPT) is a pair ((X, {Bi }); f ), where (X, {Bi })
is a partial Boolean algebra and f is a state defined on it. Equivalently, a partial
probability theory can be denoted with ((X, {Bi }); {μi }), where μi = f |B i , or simply
by ({Bi }; {μi }).
So far, such a definition just constrain PPTs to behave as classical probabilities
when restricted to contexts, and to have a noncontextual definition of their elements.
Such a definition is basically the same as that of nonsignalling theories for Bell
scenario [38], or, nondisturbing for noncontextuality scenario [39]. In their original
formulation Kochen and Specker [5, 40] proposed the following additional property
as a definition of PBA
Random documents with unrelated
content Scribd suggests to you:
Habiendo vivido santamente, murió muchos años antes que
Beatriz, la cual, antes de su muerte, escribió ella misma su vida,
como aquí se ha dicho con nombre de desengaño; porque en él ven
las damas lo que deben temer, pues por la crueldad y porfía de un
hombre padeció tantos trabajos la reina Beatriz, la que en toda Italia
es tenida por santa, donde vi su vida manuscrita, estando allá con
mis padres.
Y advierto esto, porque si alguno hubiere oído algo de esta reina,
será como digo, mas no impresa ni manoseada de otros ingenios; y
como se ha propuesto que estos desengaños han de ser sobre casos
verdaderos, fuerza es que alguno los haya oído en otras partes, mas
no como aquí va referido.
DESENGAÑO DÉCIMO.