Vous êtes sur la page 1sur 385

Cours Information Quantique 2020

Chapitre I : Algèbre Linéaire

Pr : Mostafa Mansour
Master IISPHE
Faculté des Sciences Aïn Chock-Casablanca-

Cours Information Quantique 2020 6 novembre 2020


Rappel d’algèbre linéaire Espace Vectoriel et Produit scalaire

Espace Vectoriel
Un espace vectoriel H associé aux nombres complexes C est un ensemble
d’éléments {u, v , w , ...} sur lesquelles on a défini deux opérations :
+ une opération interne de somme z = u + v
+ une opération de produit par un nombre scalaire z = a.u
+ Le produit est distributif par rapport à la
somme :a. (u + v ) = a.u + a.v , Pour tout u, v , z ∈ H et a ∈ C.

Produit scalaire
le produit scalaire sur H est une application linéaire définit par :

(u, v ) 7−→ hu, v i

+ hλu, v i = λ∗ hu, v i avec λ∈C

Espace Hilbert
Un espace vectoriel complexe de dimension muni d’un produit scalaire est
dit espace de Hilbert.
Cours Information Quantique 2020 6 novembre 2020 3 / 10
Rappel d’algèbre linéaire Notation de Dirac

 
c1
 c2 
 
 . 
+ | Ψi =   , Ce vecteur est appelée le vecteur "Ket".
 
 . 
 
 . 
ci
+ A chaque vecteur ket | Ψi on associe un nouvel être noté hΨ| qu’on
appelle vecteur "Bra".
+ hΨ| = |c1∗ , c2∗ , ..., ci∗ |.
+ L’ensemble des vecteurs "bras" constitue un espace qu’on note H∗ et
qu’on appelle espace dual de H.
+ A tout ket correspond un Bra
| ψi = λ1 | ψ1 i + λ2 | ψ2 i ⇒ hψ| = λ∗1 hψ1 | + λ∗2 hψ2 |
+ A tout Bra ne correspond pas nécessairement un ket, car de façon
générale l’espace dual H∗ de H, ne lui est pas isomorphe.

Cours Information Quantique 2020 6 novembre 2020 4 / 10


Rappel d’algèbre linéaire Notation de Dirac
Rappel d’algèbre linéaire Produit scalaire
Metric Space
Rappel d’algèbre linéaire Notation de Dirac
Vector spaces.
Rappel d’algèbre linéaire

consider again the set {|0 , |1 }.

   
1 0
|0 = , |1 =
0 1

Therefore, we have

 
1
0|0 = (1 0) =1∗1+0∗0=1
0
 
0
0|1 = (1 0) =1∗0+0∗1=0
1
 
1
1|0 = (0 1) =0∗1+1∗0=0
0
 
0
1|1 = (0 1) =0∗0+1∗1=1
1
Vector spaces.
UNIQUENESS OF A SPANNING SET
W consider the complex vector space C2, consisting of column vectors with two
elements.

         
α α 0 1 0
= + =α +β
β 0 β 0 1

And therefore the set


   
1 0
|0 = , |1 =
0 1

spans C2 , the vector space in which qubits live. Now consider the set:
   
1 1
|u1 = , |u2 =
1 −1
Rappel d'Algèbre linéaire
Projection Operators
Linear operators.
Linear operators.
Linear operators.
Linear operators.
Linear operators.
Linear operators.
Linear operators.
Linear operators.
Linear operators.
Linear operators.
Linear operators.
Linear operators.
Linear operators.
Normal Matrices.
Normal Matrices.
Normal Matrices.
Normal Matrices.
Tensor Products.
Tensor Products.
Tensor Products.
Tensor Products.
Tensor Products.
Problem 02

Solution 02
Tensor Products.
Tensor Products.
Tensor Products.
Tensor Products.
Tensor Products.
Tensor Products.
Tensor Products.
Tensor Products.
Tensor Products.
Tensor Products.
Tensor Products.
Physical Systems -
Quantum Mechanics Connections

Isolated physical 
Postulate Hilbert Space
system 

Evolution of a  Unitary
Postulate
physical system  transformation

Measurements of a  Measurement
Postulate
physical system  operators

Composite physical  Tensor product


Postulate
system  of components
The state of a quantum bit in a 2-dimensional complex Hilbert space can be
described by a unit vector (in Dirac notation)

  0   1
where α and β are complex coefficients called the amplitudes of the basis states
|0>and |1> and

   1
2 2

In conventional linear algebra

 0  1   0 
         
 1   0  1 
Global Phase

• If  '  ei  ; Im( )  0

• Then  '  '   e i ei    


2 2 2
i
u '
i
n e u 
i
n  u 
i
n

' 
• Therefore these two vectors and represent

the same physical state


(tensoriel Product)

For   
3.4

• If   1 Aˆ un  an un

• Then the state of the system    cn u n


n
• 4th postulate: The probability of measuring an
eigenvalue an of an observable A in a certain state of
the system is:

Pan   cn  un 
2 2

   cn u n    n
n n

 n  cn u n
3.4
Measurement

 n  cn un  un  un  un un   P̂n 

 n  n  cn   Pˆn † Pˆn    Pˆn Pˆn    P̂n 


2

Pan   cn  un 
2
  P̂n 
2

   cn u n    n
n n

 n  cn u n
Measurement
• The mean value of an observable: Aˆ   an P an 

  an u n 
n
  an  un un 
2

n n

   an un un     Aˆ un un 
n n

 
  Â un un     Â 
n 

Aˆ   Aˆ 

Spectral decomposition
• If   1 A uni  an uni ; i  1,2,..., g n
gn
• Then the state of the system    c u i
n
i
n
n i 1
• 4th postulate: The probability of measuring an
eigenvalue an of an observable A in a certain state of
the system is:
gn gn
P an    c  u 
2
i 2 i
n n
i 1 i 1
gn
   cni uni   n
n i 1 n
gn
n  c u i
n
i
n
i 1
Spectral decomposition
gn gn
 gn i i 
n  c u i
n
i
n  u  u i
n
i
n
   un un    Pn 
i 1 i 1  i 1 
gn
n n   c i 2
n   Pn Pn 

  Pn Pn    Pn 
i 1

gn gn
P an    c  u 
2
i 2
n
i
n   Pn 
i 1 i 1
gn
   c u i
n
i
n   n
n i 1 n
gn
n  c u i
n
i
n
i 1
Spectral decomposition
• The mean value of an observable: A    an P an 
gn gn n

  an  uni 
2
  an   uni uni 
n i 1 n i 1
gn gn
   an u i
n u     A u
i
n
i
n u 
i
n
n i 1 n i 1

 gn i i 
  A un un     A 
 n i 1 

A    A
Reduction via measurement
• When the measurement is performed only one
possible result is obtained

• Then the state of the system after the measurement


of a non-degenerate eigenvalue is:
an
  un
• For a degenerate case:

Pn 
an gn
1
  n n
c i
u i

 Pn 
gn

c
i 1
i 2 i 1
n
Exemple
Exemple
RMS deviation
• How can one quantify the dispersion of the
measurements around the mean value?

• Averaging a deviation from the average is not


adequate: A  A  A  A  0

• Instead, the RMS deviation is used:

A   A  A    A  
2 2
A

 A  2A A  A
2
 A 2 A  A
2 2 2 2

 A  A
2 2
Tensor products of state spaces
• Let there be two spaces E1 and E2 with dimensions
N1 and N2

• Tensor product of E1 and E2 is a vector field E with


the following properties:

• Notation: E  E1  E2
• If vectors belonging to E1 and E2 are  (1) ;  (2)
• Then vectors belonging to E are  (1)   (2)
• Tensor product is linear:
  (1)   (2)     (1)   (2)    (1)    (2) 
Tensor products of state spaces
• Let there be two spaces E1 and E2 with dimensions
N1 and N2

• Tensor product of E1 and E2 is a vector field E with


the following properties:

• Tensor product is distributive:


 (1)   1 (2)   2 (2)  
  (1)  1 (2)   (1)   2 (2)
• Tensor product of bases is a basis

 u (1)  E ; v (2)  E u (1)


i 1 j 2 i 
 v j (2)  E  E1  E2
Tensor products of state spaces
• If:  (1)   ai ui ;  (2)   b j v j
i j
• Then:  (1)   (2)   ai b j ui  v j
i, j

• Components of a tensor product of two vectors are


products of the components

• Not all the vectors in E can be represented as tensor


products of vectors from E1 and E2:

   ci j ui  v j ci j  ai b j
i, j
Tensor products of state spaces
• Scalar product:
  (1)   (2)   (1)   (2)    (1)  (1)  (2)  (2)
• For orthonormal bases:

 u (1)  v (2)  u (1)  v (2)    


i j k l ik jl

• Tensor product of operators:


 A(1)  B(2)  (1)   (2)   A(1)  (1)   B(2)  (2) 
• Projector:

  (1)   (2)   (1)   (2)  


   (1)  (1)     (2)  (2) 
Tensor products of state spaces
• If: A(1) m (1)  am m (1) ; B(2)  n (2)  bn  n (2)
• Then: A(1) m (1)   n (2)   am  m (1)   n (2) 
B(2) m (1)   n (2)   bn  m (1)   n (2) 
• And:
 A(1)  B(2) m (1)   n (2)  
 am  bn   m (1)   n (2) 
• Eigenvectors of A(1) + B(2) are tensor products of
eigenvectors of A(1) and eigenvectors of B(2)

• If there is one CSCO in E1 and another CSCO in E2


their tensor product is a CSCO in E
Quantum bits ( Qubits)
Bits
Bits
A bit is usually stored in a classical computer using a
transistor that is either conducting an electric signal (the
"high voltage" 1 state) or not conducting (the "low voltage"
0 state)
Qubits
Two dimensional quantum systems are called qubits

A qubit has a wave function which we write as

Examples:
Valid qubit wave functions:

Invalid qubit wave function (not normalized):


Two Level Atom as a qubit
Electron Electron on
on lower higher orbit
orbit
DC Josephson Effect B.D.Josephson Phys. Lett. 1, 251 (1962)

  x Phase difference between superconductors

qL qR
Current flow without voltage drop

Cooper pair
x dc Josephson effect
SC d SC Josephson current
I or NM I; Insulator
NM; Normal metal IJ = Ic sin(q )

Current-Voltage characteristic Thickness dependence of Ic


Penetration depth of order
I Ic parameter to barrier

I;  I  Several tens A
Josephson critical current

Ic NM ; NM  A few m

0 V 0 d 18
Transmon Qubit

01  12
12 Josephson
Energy

tunnel
1 ~ mm
01 junction
0

01 ~ 5  10 GHz

11
10 mobile electrons
Superconductivity gaps out single-particle excitations
Transmon Qubit

01  12 Anharmonicity allows us


to approximately treat
12 oscillator as a two-level ‘spin’.
Energy

1
01
0 1  

01 ~ 5  10 GHz
0  

1
2

    
Examples of coherent

 
superpositions
1
   
2
Measuring Qubits

When we observe a qubit, we get the result 0 or the result 1

0 or 1

If before we observe the qubit the wave function of the qubit is

then the probability that we observe 0 is

and the probability that we observe 1 is

“measuring in the computational basis”


Measuring Qubits Continued
new wave function
probability 0

probability

new wave function


1

The wave function is a description of our system.


When we measure the system we find the system in one state
This happens with probabilities we get from our description
Measuring Qubits
Example:
We are given a qubit with wave function

If we observe the system in the computational basis, then we


get outcome 0 with probability

and we get outcome 1 with probability:


Measuring Qubits
Example:
We are given a qubit with wave function

If we observe the system in the computational basis, then we


get outcome 0 with probability

new wave function

and we get outcome 1 with probability:

a.k.a never
Unitary Evolution for Qubits
Unitary evolution will be described by a two dimensional
unitary matrix

If initial qubit wave function is

Then this evolves to


Unitary Evolution for Qubits
Single Qubit Quantum Circuits
Circuit diagrams for evolving qubits

quantum gate

input output
qubit qubit
wave wave
function function
quantum wire
single line = qubit

time

measurement in
computational
basis
Two Qubits
Two bits can be in one of four different states
00 01 10 11

Similarly two qubits have four different states

00 01 10 11

The wave function for two qubits thus has four components:

first qubit second qubit first qubit second qubit


Two Qubits
Examples:
When Two Qubits Are Two
The wave function for two qubits has four components:

Sometimes we can write the wave function of two qubits


as the “tensor product” of two one qubit wave functions.
“separable”
Two Qubits, Separable
Example:
Two Qubits, Entangled
Example:

Assume:

Either
but this implies
contradictions
or
but this implies
So is not a separable state. It is entangled.
Measuring Two Qubits
If we measure both qubits in the computational basis, then we
get one of four outcomes: 00, 01, 10, and 11
If the wave function for the two qubits is

Probability of 00 is New wave function is


Probability of 01 is New wave function is
Probability of 10 is New wave function is
Probability of 11 is New wave function is
Two Qubits, Measuring
Example:

Probability of 00 is

Probability of 01 is

Probability of 10 is

Probability of 11 is
Two Qubit Evolutions
Rule 2: The wave function of a N dimensional quantum system
evolves in time according to a unitary matrix . If the wave
function initially is then after the evolution correspond to
the new wave function is
Two Qubit Evolutions
Manipulations of Two Qubits
Similarly, we can apply unitary operations on only one of the
qubits at a time:

first qubit second qubit

Unitary operator that acts only on the first qubit:

two dimensional
two dimensional
unitary matrix Identity matrix

Unitary operator that acts only on the second qubit:


Quantum Gates
Qubits

• A Quantum Bit (Qubit)


is a two-level quantum
|1>
system.
|0>
• We can label the
states |0> and |1>.
• In principle, this could
be any two-level
system.
Qubits
• Unlike a classical bit, which is definitely in
either state, the state of a Qubit is in general a
mix of |0> and |1>.

  c0 0  c1 1
• We assume a normalized state:

c0  c1  1
2 2
Qubits
• For convenience, we will use the matrix
representation

1  0
0    1   
 0 1
Quantum Gate
• A Quantum Logic Gate is an operation that we
perform on one or more Qubits that yields
another set of Qubits.
• We can represent them as linear operators in
the Hilbert space of the system.
Quantum gates
• The same way
classical gates manipulate only a few bits at a time,
quantum gates manipulate only a few qubits at a time
– Usually represented as unitary matrices we already saw
• Circuit representation
…boxes and different symbols
depict operations on qubits

Wires depict qubits

6/27
Pauli-X gate
• Acts on a single qubit
Dirac notation Matrix representation Circuit representation

– Acting on pure states becomes a classical NOT gate

Dirac notation…

…is obviously more convenient for calculus

8/27
Pauli-X gate
– Acting on a general qubit state

– It is its own inverse

9/27
Pauli-Y gate
• Acts on a single qubit
Dirac notation Matrix representation Circuit representation

…another gate with no classical equivalent

10/27
Hadamard gate
• Acts on a single qubit
– Corresponding to the Hadamard transform we already saw

Dirac notation Unitary matrix Circuit representation

…obviously, no classical equivalent

– One of the most important gates for quantum computing

15/27
Hadamard gate
• An interesting example

Acting on pure states…

…gives a balanced superposition…


…both states, if measured,
give either 0 or 1 with equal probability

16/27
Hadamard gate
– Applying another Hadamard gate
• to the first result

• to the second result

17/27
CNOT gate
– Example of acting on a superposition

32/27
Universal set of quantum gates
• There is more than one
universal set of gates for classical computing
• What about quantum computing,
is there a universal set of gates
to which any quantum operation possible can be reduced to?

43/27
Universal set of quantum gates
• No, but any unitary transformation
can be approximated to arbitrary accuracy
using a universal gate set
– For example (H, S, T, CNOT)

Hadamard gate Phase gate π/8 gate CNOT gate

44/27
Quantum circuits
• The same way
classical gates can be arranged to form a classical circuit,
quantum gates can be arranged to form a quantum circuit

Unlike classical circuits,


the same number of wires
is going throughout the circuit

• Quantum circuit is the most commonly used model


to describe a quantum algorithm
45/27
Chapitre 5:
Density Matrices
Density Matrix

“ When we solve a quantum mechanical problem? what we really do


is divide the universe into two part- the system in which we are
interested and rest of the universe?”

Richard P. Feynman
(Statistical Mechanics: A set of lectures; Westview press, 1972 )

When we include the part of the universe outside the system, the motivation
of using the density matrices become clear.

7/27/2021 2
For a sate
Trace of a matrix
The trace of a matrix is the sum of its diagonal elements

e.g.

Some properties:

7/27/2021 12
density operator
A density matrix, or density operator, is used in quantum
theory to describe the statistical state of a quantum system.
The formalism was introduced by John von Neumann
(according to other sources independently by Lev Landau and
Felix Bloch) in 1927. It is the quantum-mechanical analogue to
a phase-space density (probability distribution of position and
momentum) in classical statistical mechanics. The need for a
statistical description via density matrices arises because it is
not possible to describe a quantum mechanical system that
undergoes general quantum operations such as
measurement, using exclusively states represented by ket
vectors.
density operator
• Typical situations in which a density matrix is
needed include:
– A quantum system in thermal equilibrium (at
finite temperatures),
– Nonequilibrium time-evolution that starts out of a
mixed equilibrium state,
– Entanglement between two subsystems, where
each individual system must be described by a
density matrix.
Density Matrices

7/27/2021 17
Density matrices of pure states

An alternative way of representing quantum states is in terms of density


matrices (a.k.a. density operators)

18
1

7/27/2021 34
3

Diagonal elements = Populations


Off-diagonal element= Coherence

It is important to note in 2 and 3, both cases we describe a system about


which we know nothing, ie, A state of total ignorance.

7/27/2021 35
Methods to calculate the partial trace

- Partial Trace is linear map that takes bipartite states to single system states.

- We can also trace out the first system.

- We can compute the partial trace directly from the density matrix description.

7/27/2021 55
7/27/2021 56
7/27/2021 57
7/27/2021 58
7/27/2021 59
Partial Trace can be calculated in
arbitrary basis

- In a different basis

7/27/2021 60
Partial Trace can be calculated in
arbitrary basis

Which is the same as in previous


slide for other base

7/27/2021 61
Partial Trace using matrices

7/27/2021 62
Examples: Partial trace (I)

For example, it turns out that:

7/27/2021 63
Examples: Partial trace (II)
We’ve already seen this defined in the case of 2-qubit systems: discarding the
second of two qubits.

As we see in last slide, partial trace is a matrix.


How to calculate this matrix of partial trace?

7/27/2021 64
Examples: Partial trace (III): calculating
matrices of partial traces
For 2-qubit systems, the partial trace is explicitly

and

7/27/2021 65
Unitary transformations don’t change the local density matrix

- A unitary transformation on the system that is traced out does not affect
the result of the partial trace.

- I.e.

7/27/2021 66
Distant transformations don’t change
the local density matrix
- In fact, any legal quantum transformation on the traced out system, including
measurement (without communicating back the answer) does not affect the
partial trace.

- I.e.

7/27/2021 67
Chapitre 6:
Quantum Entanglement
Entanglement
Quantum entanglement is a physical
phenomenon that occurs when pairs or groups
of particles are generated or interacted in ways
such that the quantum state of each particle
cannot be described independently of the
others, even when particles are separated by a
large distance.

2
tensor product structure of the Hilbert
space that allows two or more subsystems
to share a state, which is given by the
superposition of the bases of these
Hilbert spaces. Strictly within the
framework of a pure quantum state of two
parties, the state |ψi⟩ is called separable if
we can express it as the tensor product of
the states of the subsystems
(|Ψ⟩=|Ψ1⟩⊗|Ψ2⟩ with 1 and 2 referring to
the state of the first and second
subsystems); otherwise the state is called
entangled.
3
A B
A
B

4
How do we combine two aubits?
Two separate qubits

A B

Tensor product

5
Example: constructing the standard basis for two qubits

6
Entanglement
“Total knowledge of a
composite system does not
necessarily include
maximal knowledge of all
its parts, not even when
these are fully separated
from each other and at the
moment are not
Erwin
influencing each other at Schrödinger
all.” (1935) 7
7/27/2021 8
Notation

9
10
Schmidt Decomposition Theorem

The most important theorem in bipartite quantum


systems.

11
7/27/2021 12
7/27/2021 13
7/27/2021 14
Consider a system in a quantum state |𝜓⟩

A
Schmidt
B decomposition

15
Example

16
When is a pure state entangled?

18
19
1

20
Demonstration

21
22
23
Bipartite Entanglement

24
Separability an entanglement
Let be the state of a composite quantum system
 AB S AB
Separable
m
 AB   pr  A  B , pr  0, p r  1.
r 1 r

Entangled
m
 AB
  pr   , A B
pr  0, p r  1.
r 1 r

25
Detecting entanglement

To detect the entanglement in a bipartite quantum state   H A  H B


we can use

Schmidt decomposition
PPT criterion
Fidelity
Entanglement Witness

26
Concurrence

27
28
Concurrence

29
30
Concurrence

31
Concurrence

32
Concurrence

 We can say that this state is maximaly entagled. In face, all


Bell states are maximally entangled.
33
Thus, concurrence is calculated to be

Where we have used the orthonormality of the computational basis for two
qubits such that

and the virtue for the complex numbers


34
Concurrence of a mixed states

35
36
Concurrence of mixed state

37
Concurrence of mixed state

38
Concurrence of mixed state

39
40
Concurrence of mixed state

41
Concurrence of mixed state

42
Concurrence of a mixed
state of type X

43
On considère un état bipartite de type X

Les éléments de ρAB contraints par les conditions de normalisation et de positivité


qui sont respectivement

Pour obtenir l’expression de la concurrence de Wootters d’un système


bipartite décrit par un état de type X, on calcule les valeurs propres δi de la
matrice
Nous notons que les valeurs propres δi sont les racines carrées des valeurs
propres µi du produit

7/27/2021 45
Les valeurs propres de la matrice sont obtenues par sa diagonalisation. Après un
calcul simple, nous obtenons alors

7/27/2021 46
soit la formule

Grâce à cette formule on peut déterminer la concurrence d’un état X à deux


qubits. L’utilisation de cette formule est très pratique et simple aussi bien pour les
études théoriques de l’intrication des systèmes à deux qubits que pour les
expériences.

7/27/2021 47
48
Entanglement of formation

49
50
51
Entanglement of formation
Example for maximally entangled state:

52
53
54
Negativity

55
56
57
58
Detecting entanglement
To detect the entanglement in a bipartite quantum state   H A  H B ,
we can use

Schmidt decomposition
PPT criterion
Fidelity
Entanglement Witness

59
60
The Separability Problem

Definition

This is an NP-hard problem! (Gurvits ,2003)


61
The Separability
Problem
Theorem (Stormer, 1963;Woronowcz, 1976; Peres,
1996)

62
The best-Known criterion for detection of entanglement is the
positive partial transpose (PPT) criterion. Consequently, it is
vital to notice that for a product basis , every density matrix 𝜚
can be decomposed to

With this decomposition, one can define partial transposition with


respect to each partition by

63
More specifically , if we have a state

That is,

Then the partial transpose is geven by

So we just swap the second qubits, FOR example,


if we are computing the parcial transpose of density matrix,

|01〉〈00|→|00〉〈01|,|01〉〈10|→|00〉〈11|,|01〉〈01|→|01〉〈01|
64
65
Example We know that the Bell state

Is entangled. Show this using the Peres partial


transposition condition.

66
Solution
The density operator is

The partial transpose is found by swapping the B qubits in each term.


Hence
67
68
Example 1:
In this example we consider a state that is obviously separable

Show that this is the case using the Peres partial transposition condition

Solution
The density operator is

69
The partial transpose is

In this case the two operators happen to be the same . So

70
Example2:
• Let’s consider the density of Werner state defined by :

71
With the eigenvalues :

Consedering the PPT creterion, we obtain :

72

Vous aimerez peut-être aussi