January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book

page 3

Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

Chapter 1
by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

Qubits

1.1 Introduction
A single qubit is a two-state system, such as a two-level atom. The states
(kets) |hi and |vi of the horizontal and vertical polarization of a photon can
also be considered as a two-state system. Another example is the relative
phase and intensity of a single photon in two arms of an interferometer.
The underlying Hilbert space for the qubit is C2 . An arbitrary orthonormal
basis for C2 is denoted by { |0i, |1i }, where (scalar product)

h0|0i = h1|1i = 1, h0|1i = h1|0i = 0.

Any pure quantum state |ψi (qubit) of this system can be written, up to a
phase, as a superposition (linear combination of the states)

|ψi = α|0i + β|1i , |α|2 + |β|2 = 1 , α, β ∈ C.

The classical boolean states, 0 and 1, can be represented by a fixed pair of

orthonormal states of the qubit. The standard basis in C2 is given by
   
1 0
|0i = , |1i =
0 1

and the Hadamard basis in C2 is given by

   
1 1 1 1
|0i = √ , |1i = √ .
2 1 2 −1

3
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 4

4 Problems and Solutions

Up to an overall phase an arbitrary normalized state in C2 can be written

as  iφ 
e cos(θ)
|ψi = .
sin(θ)
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

For any orthonormal basis {|0i, |1i} in C2 we have

by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

|0ih0| + |1ih1| = I2

where I2 is the 2 × 2 identity matrix. The 2 × 2 matrices

|0ih0|, |1ih1|

are projection matrices with

|0ih0|1ih1| = 02 .

Furthermore
(|0ih1| + |1ih0|)2 = |1ih1| + |0ih0| = I2 .
Given two normalized states |ψi, |φi in C2 , then 0 ≤ |hψ|φi|2 ≤ 1 provides
a probability. Let |ψi ∈ C2 and normalized. Then

ρ = |ψihψ|

is a density matrix (pure state). We have

ρ2 = |ψihψ|ψihψ| = |ψihψ| = ρ.

If the qubit represents a mixed state one uses a two-dimensional density

matrix ρ for its representation. We therefore express one qubit as

1 1
ρ= (I2 + n · σ) ≡ (I2 + n1 σ1 + n2 σ2 + n3 σ3 )
2 2

where n ∈ R3 ,
n · n ≡ n21 + n22 + n23 ≤ 1
and σ = (σ1 , σ2 , σ3 ) denote the Pauli spin matrices
     
0 1 0 −i 1 0
σ1 = , σ2 = , σ3 = .
1 0 i 0 0 −1

For pure states we have n · n = 1 and ρ = |ψihψ|. The Pauli spin matrices
are hermitian and unitary and admit the eigenvalues +1 and −1.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 5

Qubits 5

1.2 Solved Problems

Problem 1. Any normalized state (qubit) in C2 can be written as
 
α
, α, β ∈ C, |α|2 + |β|2 = 1.
β
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com
by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

Find a parameter representation (i) if the underlying field is the set of real
numbers (ii) if the underlying field is the set of complex numbers.

Solution 1. (i) Using α = cos(θ), β = sin(θ) and the identity cos2 (θ) +
sin2 (θ) ≡ 1 for all θ ∈ R we have
 
cos(θ)
.
sin(θ)

With the map θ → θ + π/2 we can construct the orthonormal basis

     
cos(θ) cos(θ + π/2) − sin(θ)
, = .
sin(θ) sin(θ + π/2) cos(θ)

(ii) We have as a representation

 
eiφ cos(θ)
sin(θ)

where θ, φ ∈ R and eiφ e−iφ = 1.

Problem 2. Consider the normalized states in C2 (θ1 , θ2 ∈ [0, 2π))

   
cos(θ1 ) cos(θ2 )
, .
sin(θ1 ) sin(θ2 )

Find the condition on θ1 and θ2 such that the vector

   
cos(θ1 ) cos(θ2 )
+
sin(θ1 ) sin(θ2 )

is normalized.

Solution 2. From the condition that the vector

 
cos(θ1 ) + cos(θ2 )
sin(θ1 ) + sin(θ2 )

is normalized we obtain (sin(θ1 ) + sin(θ2 ))2 + (cos(θ1 ) + cos(θ2 ))2 = 1. Thus

1 1
sin(θ1 ) sin(θ2 ) + cos(θ1 ) cos(θ2 ) = − ⇒ cos(θ1 − θ2 ) = − .
2 2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 6

6 Problems and Solutions

Therefore, θ1 − θ2 = 2π/3 or θ1 − θ2 = 4π/3.

Problem 3. Let { |0i, |1i } be an orthonormal basis in the Hilbert space

R2 and A := |0ih0| + |1ih1|. Consider the three cases
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

   
1 0
(i) |0i := , |1i :=
by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

0 1
   
1 1 1 1
(ii) |0i := √ , |1i := √
2 1 2 −1
   
cos(θ) sin(θ)
(iii) |0i := , |1i := .
sin(θ) − cos(θ)
Find the matrix representation of A in these bases.

Solution 3. We find
     
1 0 0 0 1 0
(i) A = + =
0 0 0 1 0 1
     
1 1 1 1 1 −1 1 0
(ii) A = + =
2 1 1 2 −1 1 0 1
sin2 (θ)
 2
  
cos (θ) cos(θ) sin(θ) − cos(θ) sin(θ)
(iii) A = +
cos(θ) sin(θ) sin2 (θ) − cos(θ) sin(θ) cos2 (θ)
 
1 0
= .
0 1
For all three cases A = I2 , where I2 is the 2 × 2 unit matrix. Obviously, the
third case contains the first two as special cases. This is the completeness
relation.

Problem 4. Let { |0i, |1i } be an orthonormal basis in the Hilbert space

C2 . The NOT operation (unitary operator) is defined as
|0i → |1i, |1i → |0i.
(i) Find the unitary operator UN OT which implements the NOT operation
with respect to the basis { |0i, |1i }.
(ii) Consider the standard basis
   
1 0
|0i = , |1i = .
0 1
Find the matrix representation of UN OT for this basis.
(iii) Consider the Hadamard basis
   
1 1 1 1
|0i = √ , |1i = √ .
2 1 2 −1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 7

Qubits 7

Find the matrix representation of UN OT for this basis.

Solution 4. (i) Obviously, UN OT = |0ih1| + |1ih0| since h0|0i = h1|1i = 1

and h0|1i = h1|0i = 0.
(ii) For the standard basis we find
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com
by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

 
0 1
UN OT = .
1 0

(iii) For the Hadamard basis we find

 
1 0
UN OT = .
0 −1

Thus we see that the respective matrix representations for the two bases
are different.

Problem 5. Let |0i, |1i be an orthonormal basis in C2 . The Walsh-

Hadamard transform is a 1-qubit operation, denoted by H, and performs
the linear transform
1 1
|0i → √ (|0i + |1i), |1i → √ (|0i − |1i).
2 2
(i) Find the unitary operator UH which implements H with respect to the
basis { |0i, |1i }.
(ii) Find the inverse of this operator.
(iii) Consider the standard basis
   
1 0
|0i = , |1i =
0 1

in C2 . Find the matrix representation of UH for this basis.

(iv) Consider the Hadamard basis
   
1 1 1 1
|0i = √ , |1i = √
2 1 2 −1

in C2 . Find the matrix representation of UH for this basis.

Solution 5. (i) Obviously,

1 1
UH = √ (|0i + |1i)h0| + √ (|0i − |1i)h1|
2 2
1 1
= √ |0i(h0| + h1|) + √ |1i(h0| − h1|).
2 2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 8

8 Problems and Solutions

−1 ∗
(ii) The operator UH is unitary and the inverse is given by UH = UH = UH ,
∗
where denotes the adjoint.
(iii) For the standard basis we find
 
1 1 1
UH = √ .
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

2 1 −1
by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

(iv) For the Hadamard basis we find

 
1 1 1
UH =√ .
2 1 −1

We see that the matrix representations for each of the two bases are the
same.

Problem 6. The Hadamard operator on one qubit can be written as

1
UH = √ ((|0i + |1i)h0| + (|0i − |1i)h1|).
2

Calculate the states UH |0i and UH |1i. Calculate UH UH .

Solution 6. We obtain the normalized states

1 1
UH |0i = √ (|0i + |1i), UH |1i = √ (|0i − |1i).
2 2

Since h0|0i = h1|1i = 1 and h0|1i = h1|0i = 0 we obtain

UH UH = |0ih0| + |1ih1| = I

where I is the identity operator (2 × 2 unit matrix).

Problem 7. Let σ1 , σ2 , σ3 be the Pauli spin matrices and I2 be the 2 × 2

unit matrix. Consider the Hilbert space C2 and the linear operator (2 × 2
matrix)  
3
1 X
Π(n) := I2 + nj σj 
2 j=1

where n := (n1 , n2 , n3 ) (nj ∈ R) is a unit vector, i.e. n21 + n22 + n23 = 1.

(i) Describe the properties of Π(n), i.e. find Π† (n), tr(Π(n)) and Π2 (n).
(ii) Find the vector (φ, θ ∈ R)
 iφ 
e cos(θ)
Π(n) .
sin(θ)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 9

Qubits 9

Solution 7. (i) For the Pauli matrices we have σ1† = σ1 , σ2† = σ2 , σ3† = σ3 .
Thus Π(n) = Π† (n). Since tr(σ1 ) = tr(σ2 ) = tr(σ3 ) = 0, tr(I2 ) = 2, and the
trace operation is linear, we obtain tr(Π(n)) = 1. Since σ12 = σ22 = σ32 = I2
and
[σ1 , σ2 ]+ = 02 , [σ2 , σ3 ]+ = 02 , [σ3 , σ1 ]+ = 02
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

where [A, B]+ := AB + BA denotes the anticommutator, the expression

by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

 2
3 3 3 3
1 X 1 1X 1 XX
Π2 (n) = I2 + nj σj  = I2 + nj σj + nj nk σj σk
4 j=1
4 2 j=1 4 j=1
k=1

simplifies to
3 3
1 1X 1X 2
Π2 (n) = I2 + nj σj + n I2 .
4 2 j=1 4 j=1 j

Using n21 + n22 + n23 = 1 we obtain Π2 (n) = Π(n). Thus Π(n) is a projection
matrix.
(ii) We find
 iφ   
e cos(θ) 1 (1 + n3 )eiφ cos(θ) + (n1 − in2 ) sin(θ)
Π(n) = .
sin(θ) 2 (n1 + in2 )eiφ cos(θ) + (1 − n3 ) sin(θ)

Problem 8. The qubit trine is defined by the following states

√ √
1 3 1 3
|ψ0 i = |0i, |ψ1 i = − |0i − |1i, |ψ2 i = − |0i + |1i
2 2 2 2
where { |0i, |1i } is an orthonormal basis. Find the probabilities

|hψ0 |ψ1 i|2 , |hψ1 |ψ2 i|2 , |hψ2 |ψ0 i|2 .

Solution 8. Using h0|0i = 1, h1|1i = 1 and h0|1i = 0 we find

1 1 1
|hψ0 |ψ1 i|2 = , |hψ1 |ψ2 i|2 = , |hψ2 |ψ0 i|2 = .
4 4 4

Problem 9. The kets |hi and |vi are states of horizontal and vertical
polarization, respectively. Consider the normalized states
1 √
|ψ1 i = − (|hi + 3|vi)
2
1 √
|ψ2 i = − (|hi − 3|vi)
2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 10

10 Problems and Solutions

|ψ3 i = |hi
1 √
|φ1 i = √ (−|hi + 2e−2πi/3 |vi)
3
1 √
|φ2 i = √ (−|hi + 2e+2πi/3 |vi)
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

3
1 √
by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

|φ3 i = √ (−|hi + 2|vi).

3
Give an interpretation of these states.

Solution 9. Since hh|hi = hv|vi = 1 and hv|hi = hh|vi = 0 we find

1 1 1
hψ1 |ψ2 i = − , hψ1 |ψ3 i = − , hψ2 |ψ3 i = − .
2 2 2
Since the solution to cos(α) = −1/2 is given by α = 120o (2π/3) or α = 240o
(4π/3) we find that the first three states |ψ1 i, |ψ2 i, |ψ3 i correspond to states
of linear polarization separated by 120o (2π/3). We find the scalar product
i
hφ1 |φ2 i = − √ .
3
The states |φ1 i and |φ2 i correspond to elliptic polarization and the third
state |φ3 i corresponds to linear polarization.

Problem 10. Let  

eiφ cos(θ)
|ψi =
sin(θ)
be a normalized state in the Hilbert space C2 , where φ, θ ∈ R. Find the
density matrix ρ := |ψihψ|, tr(ρ) and ρ2 .

Solution 10. Since

hψ| = (e−iφ cos(θ), sin(θ))

we obtain the 2 × 2 density matrix

 
cos2 (θ) eiφ sin(θ) cos(θ)
ρ = |ψihψ| = .
e−iφ sin(θ) cos(θ) sin2 (θ)

Since cos2 (θ) + sin2 (θ) = 1 we obtain that tr(ρ) = 1. With hψ|ψi = 1 we
obtain ρ2 = (|ψihψ|)2 = |ψihψ|ψihψ| = |ψihψ| = ρ. Thus we have a pure
state.

Problem 11. Given the Hamilton operator Ĥ = ~ωσ1 .

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 11

Qubits 11

(i) Find the solution

|ψ(t)i = e−iĤt/~ |ψ(t = 0)i

of the Schrödinger equation

Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

d
by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

i~ |ψi = Ĥ|ψi
dt
with the initial conditions
 
1
|ψ(t = 0)i = .
0

(ii) Find the probability |hψ(t = 0)|ψ(t)i|2 .

(iii) The solution of the Heisenberg equation of motion

dσ3
i~ = [σ3 , Ĥ](t)
dt
is given by
σ3 (t) = eiĤt/~ σ3 e−iĤt/~
where σ3 (t = 0) = σ3 . Calculate σ3 (t).
(iv) Show that hψ(t = 0)|σ3 (t)|ψ(t = 0)i = hψ(t)|σ3 |ψ(t)i.

Solution 11. (i) The solution of the Schrödinger equation is given by

|ψ(t)i = exp(−iĤt/~)|ψ(t = 0)i.

Since σ12 = I2 we find the unitary matrix

 
cos(ωt) −i sin(ωt)
exp(−iĤt/~) ≡ U (t) = .
−i sin(ωt) cos(ωt)

Thus the normalized state at time t is

   
1 cos(ωt)
|ψ(t)i = U (t) = .
0 −i sin(ωt)

(ii) We find the probability |hψ(t = 0)|ψ(t)i|2 = cos2 (ωt).

(iii) Since the commutators are given by

[σ3 , Ĥ] = ~ω[σ3 , σ1 ] = 2i~ωσ2 , [σ2 , Ĥ] = ~ω[σ2 , σ1 ] = −2i~ωσ3

we obtain the linear system of matrix-valued differential equations

dσ3 dσ2
= 2ωσ2 (t), = −2ωσ3 (t)
dt dt
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 12

12 Problems and Solutions

with the initial conditions σ3 (t = 0) = σ3 and σ2 (t = 0) = σ2 . Here we used

the Heisenberg equation of motion for σ2 to obtain the second differential
equation. The solution of this system of matrix-valued linear differential
equations is given by
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

σ3 (t) = σ3 cos(2ωt) + σ2 sin(2ωt)

by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

σ2 (t) = σ2 cos(2ωt) − σ3 sin(2ωt).

(iv) We find
hψ(t = 0)|σ3 (t)|ψ(t = 0)i = cos(2ωt)
and
hψ(t)|σ3 |ψ(t)i = cos2 (ωt) − sin2 (ωt) ≡ cos(2ωt).

Problem 12. Consider a Mach-Zehnder interferometer in which the

beam pair spans a two-dimensional Hilbert space with orthonormal basis
{ |0i, |1i }. The state vectors |0i and |1i can be considered as orthonormal
wave packets that move in two given directions defined by the geometry of
the interferometer. We may represent mirrors, beam splitters and relative
UP phase shifts by the unitary matrices
     iχ 
0 1 1 1 1 e 0
UM = , UB = √ , UP =
1 0 2 1 −1 0 1

respectively. Consider the density matrix

ρin = |0ih0|

where { |0i, |1i } denotes the standard basis. Using this basis find

ρout = UB UM UP UB ρin UB† UP† UM

†
UB† .

Give an interpretation of the result.

Solution 12. Since

   
1 1 0
ρin = |0ih0| = (1 0) =
0 0 0

and  
1 eiχ + 1 eiχ − 1
UB UM UP UB =
2 −eiχ + 1 −eiχ − 1
we obtain  
1 1 + cos(χ) i sin(χ)
ρout = .
2 −i sin(χ) 1 − cos(χ)
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 13

Qubits 13

This yields the intensity along |0i as I ∝ 1 + cos(χ). Thus the relative UP
phase χ could be observed in the output signal of the interferometer.

Problem 13. Let { |0i, |1i } be an orthonormal basis in C2 .

(i) Find the commutator [ |0ih1|, |1ih0| ].
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

(ii) Find the operators exp(t|0ih1|) and exp(t|1ih0|).

by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

(iii) Find the operator exp(t|0ih1|) exp(t|1ih0|).

(iv) Find the operator exp(t(|0ih1| + |1ih0|)).
(v) Is exp(t(|0ih1| + |1ih0|)) = exp(t|0ih1|) exp(t|1ih0|)?

Solution 13. (i) We have

h i
|0ih1| , |1ih0| = |0ih0| − |1ih1|

since h0|0i = h1|1i = 1 and h0|1i = h1|0i = 0. We see that the commutator
is nonzero.
(ii) Since h0|1i = h1|0i = 0 we find
∞ j
X t
exp(t|0ih1|) = (|0ih1|)j = I2 + t|0ih1|.
j=0
j!

Analogously
∞ j
X t
exp(t|1ih0|) = (|1ih0|)j = I2 + t|1ih0|.
j=0
j!

(iii) Multiplying the results found above we obtain

exp(t|0ih1|) exp(t|1ih0|) = I2 + t(|0ih1| + |1ih0|) + t2 |0ih0|.

(iv) Since (|0ih1| + |1ih0|)2 = I2 we obtain

∞ ∞
X t2j X t2j+1
exp(t|0ih1| + t|1ih0|) = I2 + (|0ih1| + |1ih0|)
j=0
(2j)! j=0
(2j + 1)!
= cosh(t)I2 + sinh(t)(|0ih1| + |1ih0|).

(v) Clearly when t 6= 0 we have

exp(t(|0ih1| + |1ih0|)) 6= exp(t|0ih1|) exp(t|1ih0|).

Problem 14. Consider the unitary matrix for the NOT gate
 
0 1
UN OT = .
1 0
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 14

14 Problems and Solutions

Show that we can find a unitary matrix V such that V 2 = UN OT . Thus V

would be the square root NOT gate. What are the eigenvalues of V ?

Solution 14. We find the unitary matrix

Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

 
1 1+i 1−i
V = .
by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

2 1−i 1+i

Obviously −V is also a square root. The eigenvalues of V are 1 and i. The

−V are −1√and −i. Note that the eigenvalues of UN OT are
eigenvalues of √
1 and −1 and 1 = ±1, −1 = ±i.

Problem 15. Let σ1 , σ2 , σ3 be the Pauli spin matrices. Let n be a unit

vector in R3 . We define the operator

Σ := n · σ ≡ n1 σ1 + n2 σ2 + n3 σ3 .

(i) Calculate the matrix Σ2 . From this result and the fact that Σ is hermi-
tian show that Σ is unitary. Find the eigenvalues of Σ.
(ii) Let  
1
|ψi = .
0
Calculate the state Σ|ψi and the probability |hψ|Σ|ψi|2 .

Solution 15. (i) Using n21 + n22 + n23 = 1, σ12 = σ22 = σ32 = I2 and

σ1 σ2 + σ2 σ1 = 02 , σ1 σ3 + σ3 σ1 = 02 , σ2 σ3 + σ3 σ2 = 02

we obtain

Σ2 = (n1 σ1 + n2 σ2 + n3 σ3 )2
= (n21 + n22 + n23 )I2
+n1 n2 (σ1 σ2 + σ2 σ1 ) + n1 n3 (σ3 σ1 + σ1 σ3 ) + n2 n3 (σ2 σ3 + σ3 σ2 )
= I2 .

Since Σ is hermitian, i.e. Σ = Σ∗ and Σ2 = I2 we find that Σ is a unitary

matrix with Σ = Σ−1 . Since Σ is hermitian and unitary the eigenvalues
λ1 , λ2 can only be ±1. Since tr(Σ) = 0 = λ1 + λ2 we obtain that the
eigenvalues are +1 and −1.
(ii) We find      
0 0 1
Σ|ψi = n1 + n2 + n3 .
1 i 0
It follows that |hψ|Σ|ψi|2 = n23 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 15

Qubits 15

Problem 16. Let n be a unit vector in R3 , σ = (σ1 , σ2 , σ3 ) and

n · σ := n1 σ1 + n2 σ2 + n3 σ3 .

(i) Find the unitary matrix exp(iθn · σ), where θ ∈ R.

Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

(ii) Find the normalized state

by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

 
1
exp(iθn · σ) .
0

Solution 16. (i) Since

3
X
σj σk = δjk I2 + i jk` σ`
`=1

where 123 = 231 = 312 = 1, 321 = 213 = 132 = −1 and 0 otherwise, we

obtain

exp(iθn · σ) = I2 cos(θ) + i(n · σ) sin(θ)

 
cos(θ) + in3 sin(θ) i(n1 − in2 ) sin(θ)
= .
i(n1 + in2 ) sin(θ) cos(θ) − in3 sin(θ)

Note that we could also use (n · σ)2 = I2 to find the result.

(ii) Using (i) we find the state
   
1 cos(θ) + in3 sin(θ)
exp(iθn · σ) = .
0 i(n1 + in2 ) sin(θ)

Problem 17. Consider the Hamilton operator

 
0 α
Ĥ = ~ω
α 1

where α ≥ 0. Find α where the energy gap between the two energy levels
is the smallest.

Solution 17. From the eigenvalue equation we find E 2 − ~ωE = ~2 ω 2 α2 .

Consequently
p p
E0 (α) = ~ω(1 − 1 + α2 ), E1 (α) = ~ω(1 + 1 + α2 ).

Thus p
E1 (α) − E0 (α) = 2~ω 1 + α2 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 16

16 Problems and Solutions

Therefore the shortest energy gap is for α = 0.

Problem 18. Consider the Hamilton operator

 
~ω ∆
Ĥ = ~ωσ3 + ∆σ1 =
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

∆ −~ω
by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

where ∆ ≥ 0.
(i) Find the eigenvalues and the normalized eigenvectors of Ĥ.
(ii) Use the Cayley-Hamilton theorem to calculate exp(−iĤt/~).

Solution 18. (i) From det(Ĥ − EI2 ) = 0 we find the two eigenvalues
p
E± = ± ~2 ω 2 + ∆2 .
√
We set E := ~2 ω 2 + ∆2 . Then from the eigenvalue equation
    
~ω ∆ u1 u1
= E+
∆ −~ω u2 u2

for the eigenvalue E+ = E we find ∆u2 = (E −~ω)u1 . Thus the eigenvector

is given by  
∆
.
E − ~ω
After normalization we have
 
1 ∆
.
E − ~ω
p
∆2 + (E − ~ω)2

Analogously we find for the eigenvalue E− = −E the normalized eigenvec-

tor  
1 ∆
.
∆2 + (E + ~ω)2 −E − ~ω
p

(ii) Since E+ 6= E− and E+ = E, E− = −E we have to solve the system of

equations
e−iEt/~ = c0 + c1 E, eiEt/~ = c0 − c1 E
for c0 and c1 . Then
 
−iĤt/~ c0 + c1 ~ω c1 ∆
e = c0 I2 + c1 Ĥ = .
c1 ∆ c0 − c1 ~ω

The solution of the system of equations is given by

e−iEt/~ − eiEt/~ −i sin(Et/~)

c0 = cos(Et/~), c1 = = .
2E E
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 17

Qubits 17

Thus
 
cos(Et/~) − i sin(Et/~)~ω/E −i sin(Et/~)∆/E
e−iĤt/~ = .
−i sin(Et/~)∆/E cos(Et/~) + i sin(Et/~)~ω/E

Obviously, exp(−iĤt/~) is a unitary matrix.

Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com
by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

Problem 19. Consider the Pauli spin matrices σ1 , σ2 , σ3 . Can one find
an α ∈ R such that exp(iασ3 )σ1 exp(−iασ3 ) = σ2 ?

Solution 19. We have

 
0 e2iα
exp(iασ3 )σ1 exp(−iασ3 ) = .
e−2iα 0
Thus we have to solve the equations exp(2iα) = −i, exp(−2iα) = i. For
α ∈ [0, 2π) we obtain α = 3π/4.

Problem 20. Let n and m be a unit vectors in R3 , σ = (σ1 , σ2 , σ3 ) and

n · σ := n1 σ1 + n2 σ2 + n3 σ3 .
Calculate the commutator [n · σ, m · σ].

Solution 20. We find

[n · σ, m · σ] = 2i((n2 m3 − m2 n3 )σ1 + (n3 m1 − m3 n1 )σ2
+(n1 m2 − m1 n2 )σ3 )
= 2i(n × m) · σ
where × denotes the vector product. The vector n × m is perpendicular to
the plane spanned by the vectors n and m.

Problem 21. Let |ψ1 i and |ψ2 i be two normalized states in a Hilbert
space H. A distance d with 0 ≤ d ≤ π/2 can be defined as
cos2 (d) := |hψ1 |ψ2 i|2 .
Let H = C2 and consider the normalized states
   
1 1 1 1
|ψ1 i = √ , |ψ2 i = √ .
2 1 2 −1
Find d.

Solution 21. Since hψ1 |ψ2 i = 0 we have cos2 (d) = 0 and therefore
d = π/2.
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 18

18 Problems and Solutions

Problem 22. Let ρ1 and ρ2 be density matrices in the same Hilbert

space. The Bures distance between the two density matrices is defined as
q
1/2 1/2
DB (ρ1 , ρ2 ) := 2(1 − tr((ρ1 ρ2 ρ1 )1/2 )).
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

Consider the density matrices

by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

   
1 0 1/2 0
ρ1 = , ρ2 =
0 0 0 1/2

acting in the Hilbert space C2 . Find the Bures distance.

Solution 22. Since  

1/2 1 0
ρ1 = ρ1 =
0 0
we obtain  
1/2 1/2 1/2 0
ρ1 ρ2 ρ1 = .
0 0
q √
Thus DB (ρ1 , ρ2 ) = 2(1 − 1/ 2).

Problem 23. (i) Consider the Hilbert space C2 . Show that

   
1 1 −i 1 1 i
ΠS = , ΠA =
2 i 1 2 −i 1
are projection matrices.
(ii) Decompose the Hilbert space into sub-Hilbert spaces using the result
from (i).

Solution 23. (i) We have

ΠS = Π∗S , Π2S = ΠS , ΠA = Π∗A , Π2A = ΠA

and ΠS + ΠA = I2 , ΠS ΠA = 02 .
(ii) Consider the normalized state
 iφ 
e sin(θ)
|ψi = .
cos(θ)

Then
   
eiφ sin(θ) − i cos(θ) eiφ sin(θ) + i cos(θ)
ΠS |ψi = , ΠA |ψi =
ieiφ sin(θ) + cos(θ) −ieiφ sin(θ) + cos(θ)

with hψ|ΠA ΠS |ψi = 0.

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 19

Qubits 19

Problem 24. Let σ1 , σ2 , σ3 be the Pauli spin matrices. Consider the

normalized state in C2
 iφ 
e cos(θ)
ψ= ⇒ ψ ∗ = ( e−iφ cos(θ) sin(θ) ) .
sin(θ)
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

Find the vector v = (v1 v2 v3 )T in R3 with

by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

v1 = ψ ∗ σ1 ψ, v2 = ψ ∗ σ2 ψ, v3 = ψ ∗ σ3 ψ.

Is the vector v normalized?

Solution 24. Utilizing that

eiφ + e−iφ ≡ 2 cos(φ), ieiφ − ie−iφ ≡ −2 sin(φ),

1
cos(θ) sin(θ) ≡ sin(2θ), cos2 (θ) − sin2 (θ) ≡ cos(2θ)
2
we obtain

v1 = ψ ∗ σ1 ψ = cos(φ) sin(2θ)
v2 = ψ ∗ σ2 ψ = − sin(φ) sin(2θ)
v3 = ψ ∗ σ3 ψ = cos(2θ).

The vector    
v1 cos(φ) sin(2θ)
v =  v2  =  − sin(φ) sin(2θ) 
v3 cos(2θ)
is normalized, i.e. kvk2 = 1.

Problem 25. (i) Consider the symmetric matrix over R

 
h11 h12
H=
h12 h22

and the normalized state

 
cos(θ)
|ψi = .
sin(θ)

Calculate the variance VH (|ψi) := hψ|H 2 |ψi − (hψ|H|ψi)2 .

(ii) Consider the Hadamard matrix
 
1 1 1
H=√
2 1 −1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 20

20 Problems and Solutions

and the normalized state

 
cos(θ)
|ψi = .
sin(θ)

Calculate the variance VH (ψ) := hψ|H 2 |ψi − (hψ|H|ψi)2 and discuss the
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

dependence on θ.
by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

Solution 25. (i) We find

hψ|H 2 |ψi = h211 cos2 (θ) + h222 sin2 (θ) + h212 + 2h12 (h11 + h22 ) cos(θ) sin(θ)

and

(hψ|H|ψi)2 = h211 cos4 (θ) + h222 sin4 (θ) + 2h11 h22 cos2 (θ) sin2 (θ)
+4h11 h12 cos3 (θ) sin(θ) + 4h22 h12 cos(θ) sin3 (θ).

Thus

VH (|ψi) = (h211 + h222 − 2h11 h22 ) sin2 (θ) cos2 (θ) + h212 (1 − 4 cos2 (θ) sin2 (θ))
+2h12 h11 sin(θ) cos(θ)(1 − 2 cos2 (θ))
+2h12 h22 sin(θ) cos(θ)(1 − 2 sin2 (θ)).

(ii) Using that H 2 = I2 we have

1
VH (ψ) = hψ|I2 |ψi − (hψ|H|ψi)2 = 1 − (cos(2θ) + sin(2θ))2
2
1
= (1 − sin(4θ)).
2
For θ = 0 we have VH (ψ) = 1/2. The minimum value is 0, for example for
θ = π/8. The maximum value is 1, for example for θ = 3π/8.

Problem 26. Let c† , c be Fermi creation and annihilation operators with

[c† , c]+ = c† c + cc† = I

and (c† ) = 0, c2 = 0, where I is the identity operator and 0 the zero

operator. Consider the Hamilton operator

Ĥ = ~ω1 (eiφ c† + e−iφ c) + ~ω2 c† c

and the basis |0i, c† |0i. The dual basis is h0|, h0|c. Find the matrix repre-
sentation of Ĥ and the eigenvalues of the matrix.

Solution 26. We obtain

Ĥ|0i = ~ω1 eiφ c† |0i

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 21

Qubits 21

Ĥc† |0i = ~ω1 e−iφ |0i + ~ω2 c† |0i.

Hence we obtain the matrix representation of Ĥ

~ω1 e−iφ
 
0
.
~ω1 eiφ ~ω2
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com
by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

The eigenvalues are given by

1
q
~ω2
E± = ± 4~2 ω12 + ~2 ω22 .
2 2

Problem 27. Let σ1 , σ2 , σ3 be the Pauli spin matrices and I2 the 2 × 2

identity matrix. Find the eigenvalues and normalized eigenvectors of the
Hamilton operator

Ĥ = ε0 I2 + ~ωσ3 + ∆1 σ1 + ∆2 σ2

where ε0 > 0. Are the normalized eigenvectors orthonormal to each other?

Solution 27. In matrix form we have the (hermitian) 2 × 2 matrix

 
ε0 + ~ω ∆1 − i∆2
Ĥ = .
∆1 + i∆2 ε0 − ~ω

From

det(Ĥ − EI2 ) = (ε + ~ω − E)(ε − ~ω − E) − (∆1 + i∆2 )(∆1 − i∆2 ) = 0

we obtain the characteristic equation

E 2 − 2ε0 E + ε20 = ~2 ω 2 + ∆21 + ∆22 .

Thus the two eigenvalues E+ , E− are

q
E± = ε0 ± ~2 ω 2 + ∆21 + ∆22 .
p
Let S := ~2 ω 2 + ∆21 + ∆22 . For the eigenvector of E+ we have to solve
    
ε0 + ~ω ∆1 − i∆2 v1 v1
= E+
∆1 + i∆2 ε0 − ~ω v2 v2
or

(ε0 + ~ω)v1 + (∆1 − i∆2 )v2 = E+ v1 = (ε0 + S)v1

(∆1 + i∆2 )v1 + (ε0 − ~ω)v2 = (ε0 + S)v2 .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 22

22 Problems and Solutions

We can set v1 = S + ~ω. Thus v2 = ∆1 + i∆2 . After normalizing we have

the eigenvector
 
1 S + ~ω
.
(S + ~ω)2 + ∆2 + ∆2 ∆1 + i∆2
p
1 2
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com
by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

Analogously for E− we obtain the normalized eigenvector

 
1 ∆1 − i∆2
.
(S + ~ω)2 + ∆21 + ∆22 −S − ~ω
p

Obviously the two eigenvectors are orthonormal to each other, i.e. the
scalar product vanishes. The eigenvectors do not depend on ε0 .

Problem 28. Let  and B̂ be n × n hermitian matrices. Let |ψi be a

normalized state in the Hilbert space Cn . Then we have the inequality

1
(∆Â)(∆B̂) ≥ |h[Â, B̂]i|
2
where q q
∆ := hÂ2 i − hÂi2 , ∆B̂ := hB̂ 2 i − hB̂i2
and
hÂi := hψ|Â|ψi, hB̂i := hψ|B̂|ψi.
Consider the hermitian spin- 21 matrices
     
1 0 1 1 0 −i 1 1 0
S1 = , S2 = , S3 = .
2 1 0 2 i 0 2 0 −1

Let  = S1 and B̂ = S2 . Find states |ψi such that

1
(∆Â)(∆B̂) = |h[Â, B̂]i|
2
i.e. the inequality given above should be an equality.

Solution 28. For the commutator we find [S1 , S2 ] = iS3 . Now

1
S12 = S22 = S32 = I2 .
4
We set  
c1
|ψi = , hψ| = ( c∗1 c∗2 )
c2
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 23

Qubits 23

with c1 c∗1 + c2 c∗2 = 1 (normalization). Thus we have for the right-hand side
of the equality
    
1 1 1 1 0 c1  1 2
|h[Â, B̂]i| = |hψ|iS3 |ψi = ( c∗1 c∗2 ) = |r − r22 |
2 2 4 0 −1 c2  4 1
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

where we set c1 = r1 eiφ1 , c2 = r2 eiφ2 . Now we have

by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

r
1 1
q
∆S1 = h I2 i − hS1 i2 = 1 − (c∗1 c2 + c1 c∗2 )2
4 2
r
1 1
q
∆S2 = h I2 i − hS2 i2 = 1 − (c1 c∗2 − c∗1 c2 )2 .
4 2
Thus for the left-hand side we find
1 1
q q
(∆S1 )(∆S2 ) = √ 1 − (c1 c∗2 )2 − (c∗1 c2 )2 = √ 1 − 2r12 r22 cos(2(φ1 − φ2 )).
2 2 2 2
Thus the condition from the equality is

1 1
q
√ 1 − 2r12 r22 cos(2(φ1 − φ2 )) = |r12 − r22 |.
2 2 4

Problem 29. Consider a d-dimensional Hilbert space with two orthonor-

mal bases
|b11 i, |b12 i, . . . |b1d i ∈ B1

|b21 i, |b22 i, ... |b2d i ∈ B2 .

The two bases are said to be mutually unbiased bases if

1
|hb2j |b1k i| = √
d

for all j, k = 1, . . . , d and h | i denotes the scalar product in the Hilbert

space. Consider the Hilbert space M2 (C) of 2 × 2 matrices over C, where
the scalar product is defined as

hA|Bi = tr(AB ∗ ), A, B ∈ M2 (C).

Thus d = dim(M2 (C)) = 4. The standard basis in this Hilbert space is

given by
       
1 0 0 1 0 0 0 0
E11 = , E12 = , E21 = , E22 = .
0 0 0 0 1 0 0 1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 24

24 Problems and Solutions

Let UH be the Hadamard matrix

 
1 1 1 ∗
UH = √ , UH = UH .
2 1 −1
Show that the matrices E
ejk (j, k = 1, 2)
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com
by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

ejk = UH Ejk U ∗ ,
E j, k = 1, 2
H

and the standard basis form mutually unbiased bases.

Solution 29. Straightforward calculations yield

 
e11 = UH E11 U ∗ = 1 1 1
E H
2 1 1
 
∗ 1 1 −1
E
e12 = UH E12 UH =
2 1 −1
 
e21 = UH E21 U ∗ = 1
E
1 1
H
2 −1 −1
 
∗ 1 1 −1
E
e22 = UH E22 UH = .
2 −1 1
It follows that
|tr(E1j Ee2k )| = 1 for all j, k = 1, 2.
2
Thus we have mutually unbiased bases. Apply the vec-operator to the
matrices Ejk and Eejk (j, k = 1, 2) to find mutually unbiased bases in the
4
Hilbert space C .

Problem 30. Let x0 = ct. We define a linear bijection, h, between R4

and H(2), the set of complex 2 × 2 hermitian matrices, by
 
x0 + x1 x2 − ix3
(x0 , x1 , x2 , x3 ) → .
x2 + ix3 x0 − x1
We denote the matrix on the right hand side by H.
(i) Show that the matrix can be written as a linear combination of the Pauli
spin matrices and the identity matrix I2 .
(ii) Find the inverse map.
(iii) Calculate the determinant of the 2 × 2 hermitian matrix H. Discuss.

Solution 30. (i) We have H = x0 I2 + x1 σ3 + x2 σ1 + x3 σ2 .

(ii) Consider (a, b ∈ R)
   
a c x0 + x1 x2 − ix3
= .
c∗ b x2 + ix3 x0 − x1
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 25

Qubits 25

Comparing the entries we obtain

a+b a−b c + c∗ c∗ − c
x0 = , x1 = , x2 = , x3 = .
2 2 2 2i
(iii) We obtain
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

det(H) = x20 − x21 − x22 − x23 .

by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

This is the Lorentz metric. Let U be a unitary 2 × 2 matrix. Then

det(U HU ∗ ) = det(H).

Programming Problems

Problem 1. Consider the unary gates (2 × 2 unitary matrices)

   
0 1 1 1 1
N= , H=√ ,
1 0 2 1 −1
   
1 0 1 0
V = , W =
0 eiπ/2 0 eiπ/4
and the normalized state
 
1 1
|ψi = √ .
2 1
Calculate the state N HV W |ψi and the expectation value hψ|N HV W |ψi.

Solution 1. Applying the Maxima program

/* unary.mac */
N: matrix([0,1],[1,0]);
H: matrix([1/sqrt(2),1/sqrt(2)],[1/sqrt(2),-1/\sqrt(2)]);
V: matrix([1,0],[0,exp(%i*%pi/2)]);
W: matrix([1,0],[0,exp(%i*%pi/4)]);
psi: matrix([1/sqrt(2)],[1/sqrt(2)]);
psiT: matrix([1/sqrt(2),1/sqrt(2)]);
R1: N . H . V . W; R1: ratsimp(R1);
R2: R1 . psi; R2: ratsimp(R2);
R3: psiT . R2; R3: ratsimp(R3);
we find the unitary matrix
 
1 1 −ei3π/4
N HV W = √
2 1 ei3π/4
the normalized state
 
1 1 − ei3π/4
N HV W |ψi =
2 1 + ei3π/4
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 26

26 Problems and Solutions

and
1
hψ|N HV W |ψi = √ .
2

Problem 2. Consider the Pauli spin matrix σ1 and the normalized state
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

in C2
by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

 
cos(θ)
|ψi = .
sin(θ)
Calculate the variance
Vσ1 (ψ) := hψ|σ12 |ψi − (hψ|σ1 |ψi)2
and discuss the dependence on θ

Solution 2. The following Maxima program

/* Variancesig1 */
sig1: matrix([0,1],[1,0]);
sig2: matrix([0,-%i],[%i,0]);
sig3: matrix([1,0],[0,-1]);
psi: matrix([cos(theta)],[sin(theta)]);
psiT: transpose(psi);
Vs1: psiT . (sig1 . sig1) . psi - (psiT . sig1 . psi)^2;
Vs1: trigsimp(Vs1);
D1: diff(Vs1,theta);
list: solve(D1=0,theta);
theta1: rhs(part(list,1));
theta2: rhs(part(list,2));
theta3: rhs(part(list,3));
theta4: rhs(part(list,4));
r11: subst(theta1,theta,Vs1); r12: subst(theta2,theta,Vs1);
r13: subst(theta3,theta,Vs1); r14: subst(theta4,theta,Vs1);
provides
Vσ1 (ψ) := hψ|σ12 |ψi − (hψ|σ1 |ψi)2 = 1 + 4(cos4 (θ) − cos2 (θ)).
Differentiation with respect to θ and solving the resulting equation provides
that the variance is 1 for θ = 0 and θ = π/2. The variance is 0 for θ = π/4
and θ = 3π/4.

Problem 3. Find the eigenvalues and normalized eigenvectors of the

Hamilton operator
 
Ĥ 1 1 1
K̂ = =√ .
~ω 2 1 −1

Solution 3. The Maxima program

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 27

Qubits 27

/* Hamilton1.mac */
H: matrix([1,1],[1,-1])/sqrt(2);
list: eigenvectors(H);
p1: part(list,1);
p11: part(p1,1);
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

lam1: part(p11,1);
lam2: part(p11,2);
by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

p2: part(list,2);
v1: part(p2,1); v1: part(v1,1);
v2: part(p2,2); v2: part(v2,1);
v2T: transpose(v2);
scalar: v1 . v2T;
scalar: ratsimp(scalar);
provides the eigenvalues λ1 = −1, λ2 = 1 with the corresponding (nonnor-
malized) eigenvectors
   
√ 1 √ 1
v1 = , v2 = .
− 2−1 2−1
The two eigenvectors are orthogonal to each other, i.e. scalar=0.

1.3 Supplementary Problems

Problem 1. Consider the map f : C2 → R3 defined by

 
  sin(2θ) cos(φ)
cos(θ)
f : 7→  sin(2θ) sin(φ)  .
eiφ sin(θ)
cos(2θ)
Are the vectors in C2 and R3 normalized? Consider the four normalized
vectors in C2
       
1 1 1 1 1 1 1 1
√ , √ , √ , √ .
2 1 2 −1 2 i 2 −i
Find the vectors in R3 .

Problem 2. (i) Let x1 , x2 , x3 ∈ R and σ1 , σ2 , σ3 be the Pauli spin

matrices. Show that
sin(r)
ei(x1 σ1 +x2 σ2 +x3 σ3 ) = cos(r)I2 + i (x1 σ1 + x2 σ2 + x3 σ3 )
 r 
cos(r) + ix3 sin(r)/r i(x1 − ix2 ) sin(r)/r
=
i(x1 + ix2 ) sin(r)/r cos(r) − ix3 sin(r)/r
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 28

28 Problems and Solutions

p
where r := x21 + x22 + x23 .
(ii) Let y1 , y2 , y3 ∈ R and

X := x1 σ1 + x2 σ2 + x3 σ3 , Y := y1 σ1 + y2 σ2 + y3 σ3 .
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

Consider the maps

by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

  

x1 y1
X ↔ x =  x2  , Y ↔ y =  y2  .
x3 y3

Let x · y := x1 y1 + x2 y2 + x3 y3 (scalar product). Show that

1
x·y = tr(XY ).
2
(iii) Show that
 
x2 y3 − x3 y2
i
− [X, Y ] ↔ x × y =  x3 y1 − x1 y3  .
2
x1 y2 − x2 y1

Problem 3. Let Ĥ be a 2 × 2 hermitian matrix. Consider the normalized

state  iφ 
e cos(θ)
|ψi =
sin(θ)
in the Hilbert space C2 . Assume that

hψ|Ĥ|ψi = ~ω cos(φ) sin(2θ), hψ|Ĥ 2 |ψi = ~2 ω 2 .

Reconstruct the hermitian matrix Ĥ from these three assumptions. Note

that
1
cos(θ) sin(θ) ≡ sin(2θ), eiφ = cos(φ)+i sin(φ), e−iφ = cos(φ)−i sin(φ).
2
Show that H = ~ωσ1 .

Problem 4. Let A, B be n × n matrices over C. Let v be a normalized

(column) vector in Cn . Let hAi := v∗ Av and hBi := v∗ Bv. We have the
identity

AB ≡ (A − hAiIn )(B − hBiIn ) + AhBi + BhAi − hAihBiIn .

We approximate AB as AB ≈ AhBi + BhAi − hAihBiIn .

January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 29

Qubits 29

(i) Let n = 2 and

 
1 1
A = σ1 , B = σ2 , v= √ .
2 1
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

Find AB and AhBi + BhAi − hAihBiI2 and the distance (Frobenius norm)
between the two matrices.
by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

(ii) Let n = 2 and

 
1 1
A = σ1 , B = σ2 , v = √ .
2 −1

Find AB and AhBi + BhAi − hAihBiI2 and the distance (Frobenius norm)
between the two matrices.
(iii) Consider the case
 
cos(θ)
A = σ1 , B = σ2 , v = .
sin(θ)

Find AB and AhBi + BhAi − hAihBiI2 and the distance (Frobenius norm)
between the two matrices.

Problem 5. Let α ∈ R. Show that the vectors

   
1 cosh(α) 1 sinh(α)
v1 = q , v2 = q
sinh(α) − cosh(α)
1 + 2 sinh2 (α) 1 + 2 sinh2 (α)

form an orthonormal basis in C2 . Find v1 v1T , v2 v2T , v1 v2T .

Problem 6. Let σ1 , σ2 , σ3 be the Pauli spin matrices. Show that

1 1
Π+ = (I2 + σj ), Π− = (I2 − σj )
2 2
(j = 1, 2, 3) are projection matrices. Find the vectors
   
1 1 1 1
Π+ √ , Π− √ .
2 1 2 1

Are the vectors normalized?

Problem 7. Given two arbitrary normalized states |ψi and |φi in C2 .

Find a 2 × 2 unitary matrix U such that |ψi = U |φi, i.e. U must be
expressed in terms of the components of the states |ψi and φi. Since U is
unitary we have U −1 = U ∗ .
January 24, 2018 11:57 book-9x6 10943 - Problems and Solutions in Quantum Computing and Quantum Information book
page 30

30 Problems and Solutions

Problem 8. Let A, B be 2 × 2 hermitian matrices and

 
cos(θ)
|ψi = .
sin(θ)
Problems and Solutions in Quantum Computing and Quantum Information Downloaded from www.worldscientific.com

Find the minima of the function

by 118.71.1.254 on 04/11/20. Re-use and distribution is strictly not permitted, except for Open Access articles.

f (θ) = kAB − Ahψ|B|ψi − hψ|A|ψiB + hψ|A|ψihψ|B|ψiI2 k

where k.k denotes the norm.

Problem 9. Let |0i, |1i be an orthonormal basis in C2 and z00 , z01 , z10 ,
z11 be complex numbers. Calculate

exp(z00 |0ih0| + z01 |0ih1| + z10 |1ih0| + z11 |1ih1|).

Then set

z00 = −i~ω1 , z01 = −z10 = −i~ω2 , z11 = −i~ω3 .