Department of Physics

Indian Institute of Technology, Madras

PH5842 Advanced Topics in QCQI

Tutorial 1 Quantum Channels 10 February 2025

1. Measurement and Quantum Operations

(i) Suppose we have a single-qubit system S interacting with a single-

qubit environment E through the transform,

USE = (|0⟩⟨0|)S ⊗ IE + (|1⟩⟨1|)S ⊗ (σX )E ,

where the projections act on the system and identity or the Pauli
matrix act on the environment. Assuming the environment starts in
the state |0⟩, obtain the Kraus operators of the quantum operation on
the system for this process.
(ii) Repeat the exercise for the unitary process,
   
σX σY
USE = √ ⊗ (I)E + √ ⊗ (σX )E
2 S 2 S

(iii) More generally, suppose the environment is measured in the basis

{|ei ⟩} after the joint system-environment unitary USE has been ap-
plied. Write down the Kraus operators {Ek } of the corresponding
quantum channel induced on the system, in terms of USE and the
basis {|ei ⟩}.
Show that, if the system started out in state ρ, the probability of ob-
taining outcome k (after the action of USE followed by a measurement
in E) is given by pk = Tr(Ek ρEk† ), and the system state ρk correspond-
ing to outcome k being realized is ρk ∝ Ek ρEk† .

2. Bit-flip Channel: The action of a single-qubit bit-flip channel is described

by the quantum operation

E(ρ) = (1 − p)ρ + pσX ρσX .

between a pure state |ψ⟩ and a mixed state ρ is given by

(i) The fidelity p
F (|ψ⟩, ρ) = ⟨ψ|ρ|ψ⟩. Obtain an expression for the fidelity between
an initial pure state |ψ⟩ and the final state after the action of the
bit-flip channel. Hence evaluate the minimum fidelity (obtained by
minimizing the fidelity function over all pure, initial states) after the
action of the bit-flip channel.
 (ii) Show that the bit-flip channel may be given an alternate operator-sum
representation of the form E(ρ) = (1 − 2p)ρ + 2pP+ ρP+ + 2pP− ρP− ,
where P+ = |+⟩⟨+| and P− = |−⟩⟨−| are the projectors onto the +1
and −1 eigenstates of σX .

3. Unital maps: Recall that the action of a completely positive trace-preserving

(CPTP) map (also called a quantum channel on any state ρ is described by,
N
X
E(ρ) = Ei ρEi† ,
i=1

wherein the Kraus operators of the map {Ei } satisfy

X †
Ei Ei = I.
i

A CPTP map is said to be unital if it leaves the identity operator invariant

(E(I) = I), or equivalently,
X
Ei Ei† = I.
i

(i) Show that the bit-flip channel, phase-flip and the depolarizing channel
are unital.
(ii) Consider a pair of probability distributions P ≡ {p(i)}, Q ≡ {q(j)}.
If there exists a doubly stochastic matrix D such that P = DQ , then,
P is said to be majorized by Q; this is denoted as P ≺ Q. A doubly
stochastic matrix is a matrix with P real, non-negativePentries whose
rows and columns sum to one: i Dij = 1, ∀j and j Dij = 1 ∀i.
Such a majorization relation indicates as to which distribution is more
mixed: if P is majorized by Q, then P is more mixed that Q, that is,
the corresponding Shannon entropies satisfy H(P ) ≥ H(Q).
Now, consider two quantum states ρ, σ. We know that their corre-
sponding eigenvalue spectra Λρ ≡ {λρ (i)}, Λσ ≡ {λσ (j)} can be re-
garded as probability distributions. We say ρ ≺ σ (ρ is majorized by
σ) if the corresponding eigenvalue distributions satisfy Λρ ≺ Λσ . Show
that, if E is a unital channel, the final state after the action of a unital
map is majorized by the initial state:

E(ρ) ≺ ρ.

Note: This result shows that unital maps are the natural quantum
generalizations of doubly stochastic maps. The final state after the
action of a unital map is no less random than the original state.
 4. Single-qubit channels: Consider a single-qubit density operator ρ with
corresponding Bloch vector ⃗r ≡ (rx , ry , rz ). Recall that the phase-flip chan-
nel EP and the depolarizing channel ED are described by the following Kraus
operators:
√ √ √ r
√ p p p p 3p
EB ≡ { pσZ , 1 − pI}, ED ≡ { σX , σy , σZ , 1 − I}.
2 2 2 4

For each of these channels obtain the Bloch vector r⃗′ ≡ (rx′ , ry′ , rz′ ) of the
state ρ′ = E(ρ) = i Ei ρEi† , after a single application of the channel. Hence
P
represent pictorially, the effect of these channels on the Bloch sphere.

5. Amplitude Damping Channel: The quantum amplitude damping chan-

nel is defined by two Kraus operators, given by,
   √ 
1 √ 0 0 p
E0 = , E1 = .
0 1−p 0 0

In this exercise, we will work out how such a noise channel can arise out of
a physical system-bath interaction.

(i) Check that the channel described by the two Kraus operators above
is trace preserving. Is it also unital?
(ii) Spontaneous emission as amplitude damping: The dynamics of
an idealized two-level atom in a cavity can be modeled via the stan-
dard Jaynes-Cummings Hamiltonian which describes the interaction
between a two-level atom in a cavity and a quantized mode of the
electromagnetic field. Specifically,

Htot = Hatom + Hfield + Hint , (1)

where, Hatom = ℏω2 0 σZ is the Hamiltonian of the atom in the two-

level {|0⟩atom , |1⟩atom } subspace, and, Hfield = ℏωa† a, with a† , a being
the creation and annihilation operators of the single-mode field1 . The
interaction Hamiltonian is given (under the rotating wave approxima-
tion) as,
Hint = g(a† σ− + aσ+ ),
where σ± = σX ±iσ
2
Y
are the Pauli raising and lowering operators, and
g is the coupling constant for the interaction between the radiation
and the field.
1
The creation and annihilation operators are √ defined by their action √
in the photon-number
basis (also known as the Fock basis): a† |n⟩ = n + 1|n + 1⟩, a|n⟩ = n|n − 1⟩, where, for
n = 0, 1, 2, . . ., |n⟩ is an n−photon energy eigenstate.
 First, show that the Hamiltionian in Eq. (1) can be rewritten in terms
†
of the number operator N = a a+σ2
Z
, as,

Htot = ℏwN + δσZ + g(a† σ− + aσ+ ),

where δ = (w02−w) is known as the detuning – the frequency difference

between the field and atomic resonance.
Next, suppose we focus on the case where there is at most one exci-
tation (i.e., one photon) in the field mode of the cavity. Write down
the matrix form of the atom-field Hamiltonian in the four-dimensional
space spanned by {|0⟩field |0⟩atom , |0⟩field |1⟩field , |1⟩field |0⟩atom , |1⟩field |1⟩atom }.
Hence, show that the time-evolution operator for the atom-field system
is given by,
 
−iHt −iδt δ
U (t) = e = e |00⟩⟨00| + cos Ωt + i sin Ωt |01⟩⟨01|
Ω
 
δ
+ cos Ωt − i sin Ωt |10⟩⟨10|
Ω
g
− i sin Ωt (|01⟩⟨10| + |10⟩⟨01|) , (2)
Ω
p
where Ω ≡ g 2 + δ 2 is called the Rabi frequency.

Finally, show that the quantum operation resulting from tracing out
the radiation field mode at zero-detuning (δ = 0) is indeed an ampli-
tude damping channel.
(iii) Circuit Model: In the figure below, the first wire represents a ‘sys-
tem’ (S) qubit and the second wire represents the ’environment’ (E).
The environment qubit starts out in |0⟩ and is measured in the {|0⟩, |1⟩}
in the end. The joint system-environment (SE) unitary is described
by a sequence of controlled-unitaries as specified in the figure. Work
ρin • ρout
|0⟩ RY (θ) •

out the Kraus operators Ei such that, ρout = i Ei ρin Ei† . Hence show
P
that this is a circuit model of the amplitude damping channel.