Tutorial on Quantum Repeaters

Rodney Van Meter & Tracy Northup

@IRTF, 2019/3/25
Outline

• What’s a quantum network? (tl;dr version)

• Applications: Why quantum networks?
• Basic terminology & concepts
• Basic mathematical notation
• Basics of entanglement
• Teleportation
• Implementation
• Repeaters: rationale, concepts & generations
• Final thoughts: learning more & getting involved
2
Quantum networks: the vision

• Quantum nodes at which information

is stored and processed.
» atoms

• Quantum channels for information

transport.
» photons

H. J. Kimble, Nature 453, 1023 (2008)

3
 Two kinds of quantum networks

Unentangled Networks Entangled Networks

Good only for quantum key Good for many purposes:
distribution (QKD), which aids • crypto functions including QKD
longevity of secrecy of • high-precision sensor networks
encrypted information on Actually a series
• connecting quantum computers
of steps from here
classical networks. into a Quantum Internet.
to there (Wehner,
Very limited distance (but Unlimited distance using quantum
Elkouss, Hanson)
satellite possible!). repeaters.
Weak in multi-hop settings, Strong in networked settings.
better for point-to-point. Hard to build.
Easier (still not easy) to build.

4
Entanglement （量子もつれ）

Even if they
are far apart!

“Measure” this and you’ll also

one and find know what this
its value… one is
 The job of a quantum
repeater network is…

• …to make end-to-end entanglement (modulo some

arguments about temporal matters).

• And, entanglement is a consumable resource, so

we have to make lots of it.
 So the job of a quantum
repeater is…
• 1) to make base-level entanglement over a physical link

• 2) to couple entangled links along an end-to-end path to

meet the applications’ needs

• 3) to monitor and manage errors

(purification, QEC, or both)

• 4) to participate in the management of the network

 And the job of a Quantum
Internet is…
To do all of this:

• across heterogeneous networks (both physically and

logically)

• in an environment with minimal trust between networks

• no knowledge of the internals of autonomous

networks

• possible presence of malicious nodes

 Link level

End-to-end

11
http://www.nature.com/articles/srep20463
Good for & not good for

Quantum networks are about new capabilities, not

some path to huge communication bandwidth.
Reduced # of communication rounds
(asymptotically, theoretically), higher precision,
scalability of distributed quantum systems, etc.

No faster-than-light communication!

You can each get shared, secret

random numbers upon measuring
shared, entangled states, but that
doesn’t give you the ability to send
messages.

12
 Why quantum networks?
Introduction to applications
Reduce dependency on
public key, one-way
functions, computational Byzantine
complexity agreement
Distributed
Leader election
crypto functions

Quantum
Quantum key
secret sharing
distribution (QKD)
Blind quantum
computation
Basic Interferometry
client-server
QC Clocks
Distributed System-area
Other reference
computation networks Sensors
frame uses 14
 Byzantine
agreement
Distributed
Leader election
Low crypto functions
bandwidth
Quantum
Quantum key
secret sharing
distribution (QKD)
Blind quantum
computation
High to very High to very
Interferometry
Basic
high high
client-server
bandwidth
QC bandwidth
Clocks
Distributed System-area
Other reference
computation networks Sensors
frame uses 15
 IPsec with QKD: Quantum-protected
campus-to-campus connection
Dark Fiber
IPsec Gateway
Quantum Path for
Key Exchange

Internet

IPsec Gateway
IPsec Tunnel

Classical internet path

using Quantum-made key

draft-nagayama-ipsecme-ike-with-qkd-01.txt, 2014/10

12 16
IPsec with QKD

IP
IPsec Network IPsec
Gateway Encrypted Connection Gateway

key Secure, Local Secure, Local key

Connection Connection

Quantum
QKD Network QKD
Device Make keys Device

13 17
 TLS with QKD:
Quantum-protected web/e-commerce
Many connections
One connection

18
D-H: Diffie-Hellman key exchange
QKD: Quantum Key Distribution testing all keys
AES: Advanced Encryption Standard becomes possible
OTP: One Time Pad Factoring AES broken
becomes
possible
Data
encrypted Is this gap
today interesting?
D-H +
AES

QKD +
AES

QKD +
super-AES

QKD +
OTP

Feasible today See WeakDH.org!

for metro nets (Am interested in your opinion of this!)
Blind Computation: Secure
Quantum
Time-Sharing
Q
Q
Q Server learns nothing about either client’s data or computation,
except upper bound on computation size. (cf. classical
homomorphic encryption) (Broadbent et al.)
Distributed QC: blind computing

21
Sensors: Interferometry

DOI: http://dx.doi.org/10.1103/PhysRevLett.109.070503 22
Oh, yeah, and
communication complexity

For some abstract tasks, theoretically can be

exponentially fewer rounds of communication. I don’t
know much about this, but see Raz STOC 1999;
arXiv:quant-ph/0101005; arXiv:quant-ph/0611209;
arXiv:1605.07372; and a few others.
23
 Technical demands
Some of these require only the ability to measure an
inbound photon at the end nodes, others require
ability to entangle your local memory at the few-
qubit level. True distributed computing (e.g. blind)
requires lots of high-fidelity memory.

Wehner, Elkouss, Hanson, Science, 2018 rdv et al., arXiv:1701.04586

Basic terminology & concepts
Seven Key Concepts for Quantum Computing

• Superposition

• Interference

• Entanglement

• Unitary (reversible) operation

• Measurement

• No-cloning theorem

• Decoherence

26
(Abbreviated) Glossary

• Quantum amplitude: represented by a complex number, the “amount” of the

quantum wave function in a particular state

• Pure state: A quantum state whose preparation process did exactly what it
was supposed to (no noise, no errors, no decay). n.b.: might be in
superposition, might be entangled. Fidelity = 1.0.

• Mixed state: A quantum state with noise, errors, decay. Fidelity < 1.0.

• Entangled state: A multi-qubit quantum state whose qubits can’t be described

independently, only in the context of all the qubits.

• Bell pair: A canonical two-qubit entangled state. There are four types, can be
interconverted, and also can be used as a basis set for describing two-qubit
states.

27
Superposition: Quanta Behaving
Like Waves

+ =
Interference ／干渉／ การทับซ ้อน

Destructive
Interference

Constructive
interference

Waves are Waves are

in phase out of phase
Basic mathematical notation
(see extensive appendix for
more)
Dirac’s Bra-ket Notation

ket
Dirac’s Bra-ket Notation
ket
Dirac’s Bra-ket Notation

bra ket
 State Vector for Two & Three Qubits

There are 2^n elements in the state

vector for n qubits.
Each amplitude is a complex number.
Prob. of measuring i is
Normalization requires
34
No-cloning theorem
• In general, independent (unentangled) copy
of a quantum state cannot be made.
• This theorem is important for cryptographic
communication with quantum computation.
Basics of entanglement:
Bell pairs & nonlocality
Nonlocality - the arguments
Einstein
Either quantum mechanics is
nonlocal, or it is incomplete
(secret plans)

J.S.Bell Even if it is incomplete, it is still

nonlocal!

Aspect

No local hidden variable theory can

explain my experiment.
The EPR argument

Einstein, Podolsky, and Rosen asked: is quantum

mechanics complete? Before we measure a
system, does it have a definite state?

“If, without in any way disturbing a system,

we can predict with certainty (i.e., with
probability equal to unity) the value of a
physical quantity, then there exists an
element of physical reality corresponding
to this physical quantity.”
The EPR argument

Preconditions of the argument:

1. Locality: no immediate signaling.

2. Each measurement generates a single

result.

QM is either nonlocal OR incomplete.

Bell’s arguments

Bell wondered if adding in hidden variables really

saves QM from being nonlocal.

Bell derived an inequality that any local hidden

variable theory must satisfy.

He then showed that there are some quantum

states that can violate this inequality.

So quantum mechanics cannot be described with

a local hidden variables theory.
That’s all well and good in theory, but…

Experimentally, Bell inequality violations have

been measured convincingly in many settings.

Since 2015: In three experiments, in ways that

close all loopholes that (most) people take
seriously…

Note: There are many Bell-type inequalities.

Most famous: Clauser-Horne-Shimony-Holt (CHSH).
The Bell states

Bell states are entangled states.

That is, they can’t be written as separable states
(“product states”).
(These four can also be used as a basis set to
rewrite any two-qubit state.)
42
Bell states are a resource for quantum
communication and computation.

Bell pairs can be used to generate secret keys.

If two people share a Bell pair,

one can send a quantum state to the other.
43
The no-signaling theorem

Can we use quantum nonlocality to send a signal

faster than light?

NO - it is a mathematical consequence of
quantum mechanics that even if it is nonlocal,
quantum systems cannot signal each other.
Consequences for quantum computing

INFORMATION CANNOT TRAVEL FASTER

THAN THE SPEED OF LIGHT
(probably)

Entangled states can appear to be signaling each

other, but really they cannot.

Always look for the classical communication

channel...
Teleportation
Teleportation

•Yet another concept discovered by Charles

Bennett & co.

• Moves a quantum state from one location to

another, not the physical carrier of the qubit.
Teleportation

•Yet another concept discovered by Charles

Bennett & co.

• Moves a quantum state from one location to

another, not the physical carrier of the qubit.
Alice Bob

D D
So, how do we do this?
The Bell Basis

Stabilizers
(won’t be discussed in
this tutorial; unlikely to
come up, but if they do,
ask us later)
Our Data Qubit
Single-qubit state One-qubit operations

Measure 0 w/ prob.

What happens to our

state after applying
those ops:
Teleportation Operations

Bell pair creation

Bell State Measurement
AB
0 1
Z? X?
D D

Alice Bob
Implementation
How can I make my own Bell pair?

Need: two-level quantum systems in which to encode

information.

Wish list:
• easy to entangle with each other
• easy to measure (in different bases)
• both of those processes: fast & accurate
• minimal information loss (through interaction with the
environment)
• transportable
• identical

53
Spoiler alert: there’s no perfect TLS.

…so the best choice depends on the application.

Some favorites:
• photons
– we have: lasers, optical fibers, wave plates & detectors…
– how do you get just one? what if you want it to stay in one place?
• atoms
– controlled interactions with lasers & microwaves; storage & processing
– require sophisticated laboratories; not going anywhere
• artificial atoms
– properties can be tailored; scalable fabrication
– are they really identical? are there only two levels?
– also not going anywhere…
54
Quantum networks: the vision

• Quantum nodes at which information

is stored and processed.
» atoms

• Quantum channels for information

transport.
» photons

H. J. Kimble, Nature 453, 1023 (2008)

55
Photons

encode 0 and 1 in…

• polarization
• time bin
• number
• path

J. L. O’Brien, Science 318, 1567 (2007)

56
Atoms

identify two electronic states!

• how long do they live?
• are they addressable with lasers? microwaves?
• are they sensitive to environmental fluctuations?

57
Generating remote entanglement, probabilistically

• Each atom emits a photon (or not).

• The atom’s state depends on
whether or not it emitted a photon
(a), or the photon polarization (b).
• The detectors can’t tell which atom
the photon(s) came from.
• Detection projects the atoms into an
entangled state.

T. Northup and R. Blatt, Nature Photon. 8, 356 (2014) 58

 Entangling remote 171Yb+ ions
Monroe group, University of Maryland / JQI
B. B. Blinov, Nature 428, 153 (2004)
D. L. Moehring et al., Nature 449, 68 (2007)
D. Hucul, Nat. Phys. 11, 37 (2015)

http://iontrap.umd.edu/research/ion-photon-quantum-networks/
Entangling remote 171Yb+ ions
Monroe group, University of Maryland / JQI
B. B. Blinov, Nature 428, 153 (2004)
D. L. Moehring et al., Nature 449, 68 (2007)
D. Hucul, Nat. Phys. 11, 37 (2015)
Entangling remote 171Yb+ ions
Monroe group, University of Maryland / JQI
B. B. Blinov, Nature 428, 153 (2004)
D. L. Moehring et al., Nature 449, 68 (2007)
D. Hucul, Nat. Phys. 11, 37 (2015)
Entangling remote 171Yb+ ions
Monroe group, University of Maryland / JQI
B. B. Blinov, Nature 428, 153 (2004)
D. L. Moehring et al., Nature 449, 68 (2007)
D. Hucul, Nat. Phys. 11, 37 (2015)
Entangling remote 171Yb+ ions
Monroe group, University of Maryland / JQI
B. B. Blinov, Nature 428, 153 (2004)
D. L. Moehring et al., Nature 449, 68 (2007)
D. Hucul, Nat. Phys. 11, 37 (2015)
Entangling remote 171Yb+ ions
Monroe group, University of Maryland / JQI
B. B. Blinov, Nature 428, 153 (2004)
D. L. Moehring et al., Nature 449, 68 (2007)
D. Hucul, Nat. Phys. 11, 37 (2015)
 state fidelity

Entangling remote 171Yb+ ions

Monroe group, University of Maryland / JQI here: fidelity = 78(3)%
B. B. Blinov, Nature 428, 153 (2004)
D. L. Moehring et al., Nature 449, 68 (2007)
(classical bound: 50%)
D. Hucul, Nat. Phys. 11, 37 (2015)
entanglement every 0.22 s
Remote entanglement has been shown
in a handful of experimental systems

• atomic ensembles
C. W. Chou et al., Nature 438, 828 (2005)
• neutral atoms
J. Hofmann et al., Science 337, 72 (2012)
• NV centers
H. Bernien et al., Nature 497, 86 (2013)
• superconducting qubits
A. Narla et al., Phys. Rev. X 6, 031036 (2016)
• quantum dots
A. Delteil et al., Nat. Phys. 12, 218 (2016)

state of the art: two-node experiments

 Repeaters:
rationale, concepts &
generations
 So the job of a quantum
repeater is…
• 1) to make base-level entanglement over a link

• 2) to couple entangled links along an end-to-end

path to meet the applications’ needs

• 3) to monitor and manage errors

(purification, QEC, or both)

• 4) to participate in the management of the network

Conceptual Hardware

(There are also all-optical approaches, with no

static buffer memory.)
69
Store-and-Forward?

70
Direct Transmission Pretty Clearly Doesn’t
Work...

Loss in channel
always too high

Must use
acknowledged
link layer, build
generic Bell pair,
then teleport

71
Timing Trapezoids

72
So Why Doesn’t Hop-by-Hop Teleportation
Work?
Long memory times, swapping (local xfer) fidelity

Star: E2E teleportation; filled circle: QKD; open circle: qubit measurement timing
blue bar: entanglement swapping 73
 Link level

End-to-end

74
http://www.nature.com/articles/srep20463
Five Repeater Schemes

• 1G: Purify and swap over ACKed links: truly a distributed computation
(Dur & Briegel, Lukin, others; since 1998)
• 2G: Error Correction over ACKed links
• CSS quantum error correction & entanglement swapping
(Jiang (Lukin) et al., 2009)
• Surface code quantum error correction, sort of but not quite swap
(Fowler et al., 2010)
• 3G: Error Correction over no-ACK-needed links: store-and-forward
• Quasi-asynchronous
(Munro et al., 2010)
• Memoryless
(Munro et al., 2012)

75
 Quantum Repeater Operation

Station 0 Station 1 Station 2

Bell State
Measurement
Called entanglement swapping.
Fidelity declines; you must purify afterwards

76
Nested Entanglement Swapping

Dur & Briegel, many others 77

 Purification: Error Detection

Station 0 Station 2

Two entangled states used:

one used as a test tool to test an assertion about the other.
The test tool is destroyed in the process.
On success, confidence in the tested state (fidelity) improves.
On failure, tested state is discarded. 78
Repeater Protocol Stack

Application
Purification Control (PC) } End-to-End

Repeated at
Entang. Swapping Ctl (ESC)

Purification Control (PC)

} Different
Distances
Entanglement Control (EC)
Physical Entanglement (PE) } Distance=1
Only quantum!
Van Meter et al., IEEE/ACM Trans. on Networking,
Jun. 2009, quant-ph:0705.4128
79
Four-Hop Protocol Interactions

App App
PC PC
ESC ESC ESC
PC PC PC
ESC ESC ESC ESC ESC
PC PC PC PC PC
EC EC EC EC EC
PE PE PE PE PE

80
2G & 3G are still far away

• Even 1G is really hard

• Entanglement success probability is low, many round trips in protocols,
few qubits per node, memory lifetimes are still problematic
• Getting & keeping a Bell pair to the left at the same time as Bell pair to the
right to enable swapping
• Getting & keeping two Bell pairs for purification
• Gate errors in both purification and swapping
• 2G: Error Correction over ACKed links
• Will blow up resource requirements at least 7x, before the probabilistic
problems above
• Gate error rates too high for QEC to work yet
• 3G: Error Correction over no-ACK-needed links (store-and-forward)
• Prob. 80-93% or better (depending on code) of correctly receiving each
individual photon
81
What would the simplest 1G repeater look like?
H.-J. Briegel, W. Dür, J. I. Cirac, and P. Zoller, Phys. Rev. Lett. 81, 5932 (1998)

1. Entangle one pair of atoms.

2. Entangle a second pair of atoms.
3. A Bell measurement on the two central atoms entangles the two outermost
atoms.
This approach is not scalable because errors accumulate.

Bell measurement

entangled!
What would the simplest 1G repeater look like?
H.-J. Briegel, W. Dür, J. I. Cirac, and P. Zoller, Phys. Rev. Lett. 81, 5932 (1998)

Entanglement purification is the key to scalability.

Multiple copies of entangled pairs are reduced to fewer copies with higher
fidelity. Requirements: gate operations between local qubits & readout of one
qubit.
Quantum repeater = entanglement swapping + purification
Note that we have to wait for classical information to travel from A to B.

Bell measurement

entangled!
No one has built a quantum repeater yet.
The simplest version:
1. Eight qubits, three nodes
2. Gate operations between qubits for Bell-state measurements and
purification.
Closest experiment: entanglement purification with four qubits (two NV
centers, two nuclear spins).
N. Kalb et al., Science 356, 928 (2017)
State of research
 Some networking results from rdv’s group
Routing Multiplexing

rdv et al.,
Aparicio & rdv, SPIE, 2011
Net. Sci. 2013

Architecture Layered Protocols

rdv et al.,Trans. Networking, 2009

rdv et al.,Prog. Info. 2013
 Some networking results from rdv’s group
Internetworking Hijacking of a Repeater

Satoh et al., QST, 2018

Nagayama et al., Phys. Rev. A 2016

Network Coding Byzantine

agreement

Matsuo et al.,Phys. Rev. A 2018 Taherkhani et al.,QST, 2017

Stephanie’s group
Quantum internet: A Vision for the road ahead (https://doi.org/10.1126/science.aam9288):
This was presented already in Bangkok. This paper defines stages of quantum networks and what applications
can be realized on these.

A Link Layer Protocol for Quantum Networks (will be on arXiv tomorrow!):

Presents a link layer protocol which provides the defines service in https://datatracker.ietf.org/doc/draft-
dahlberg-ll-quantum/ together with simulations of performance metrics of a complete implementation of the
protocol using a discrete-event simulator.

Parameter regimes for a single sequential quantum repeater (https://arxiv.org/abs/1705.00043):

Assesses the performance of single sequential quantum repeaters using realistic hardware parameters.

Fully device-independent conference key agreement (https://arxiv.org/abs/1708.00798):

Presents security proof for fully device-independent conference key, which task is to distribute a secret key
among N parties.

Anonymous transmission in a noisy quantum network using the W state (https://arxiv.org/abs/1806.10973):

Shows how one can perform anonymous transmission using a W state, which in many regimes can tolerate more
noise than the more common approach of using a GHZ state.

SimulaQron – A simulator for developing quantum internet software (https://arxiv.org/abs/1712.08032):

Presenting SimulaQron, a simulator intended to be used for development of software for quantum networks.
88
 Final thoughts:
References, learning more &
getting involved
 Good Repeater References
• Dur & Briegel’s originals, 1990s onwards

• Kimble’s “Quantum Internet” in Nature, 2008

• Rodney Van Meter, Quantum Networking, Wiley-iSTE 2014

• Takeoka, Guha, Wilde, Nature Communications, 2014

on when repeaters are more effective than simple transmission

• Muralidharan et al., Scientific Reports, 2016

defines 1G, 2G, 3G

• Pirandola, Nature Communications, 2017

extending TGW analysis

• Wehner, Elkouss, Hanson, Science, 2018

roadmap

90
4th run of MOOC
(massive open online
course) on Quantum
Computing includes 日本 4th run of MOOC will
語の字幕！Starts April 1! include translation into
Thai! 92
https://www.edx.org/course/quantum-information-science-i
 We have created a Research Group
(RG) on Quantum Internet inside the
Internet Research Task Force (IRTF).
Co-chairs are Van Meter (Keio) and
Stephanie Wehner (TU Delft).
https://www.irtf.org/mailman/listinfo/qirg
95

236 list members (as of 2019/3/18)

n.b.: Dates still tentative

96
Done!
Any questions you haven’t asked yet?
Appendices

• Vectors, Matrices, Dirac’s bra-ket notation, and Complex Numbers

• Quantum States
• Single-Qubit Gates
• Two-Qubit Gates
• Density Matrix
• Teleportation by the numbers
 Vectors, Matrices,
Dirac’s bra-ket notation, and
some complex numbers
Vectors
A vector has a direction and a distance. It can be
described via a set of coordinates, assuming the
vector starts at the origin.

1.0

1.0
Dirac’s Bra-ket Notation

ket
Dirac’s Bra-ket Notation
ket
Dirac’s Bra-ket Notation

bra ket
Vector Addition
Vector Inner Product
Matrix

n.b.: It’s more

standard to start
index at 1, but 0 will
be more convenient
here.
Matrix Addition
Multiplying a vector by a matrix
 Complex numbers
Quantum physics contains complex numbers so
that
Quantum information is consist of complex
numbers.

real part imaginary part

Complex plane
Im

Euler's formula
Re
 Exponentiating
Im complex numbers

Euler's formula
Re
Complex Numbers in Bra-ket

Earlier presentation wasn’t quite complete!

complex conjugate
Vector Outer Product

Same thing as tensor!

...what’s a tensor?
Tensor Product
Tensor Product
Quantum States
Writing quantum states

To write a quantum state we use a ket vector:

A
This is “the state of system A”.

We can put many things inside kets...

Writing quantum states

To write a quantum state we use a ket vector:

We can put many things inside kets...

Writing quantum states

To write a quantum state we use a ket vector:

We can put many things inside kets...

Writing quantum states

To write a quantum state we use a ket vector:

We can put many things inside kets...

Writing quantum states

To write a quantum state we use a ket vector:

We can put many things inside kets...

Writing quantum states

Ket vectors tell us the probability of measuring the

state.

Suppose we just have two outcomes: 0 and 1.

A ket vector for the system could be written

What does this mean?

Bloch Sphere
is up here

Question: Where are

Note: This is not (usually)

directly relatable to something is down here
physical! Image from Wikipedia 123
Bloch Sphere

124
Phase

Phase, then, is position about the vertical axis

on the Bloch sphere.
You can’t touch it, see it, smell it, taste it,
or hear it.
It doesn’t “mean” anything.

...but you can calculate using it.

125
The Z Gate

Question: What does this do to this state?

Question: What does this do to any observation or

measurement we make?

126
Multiple Qubits: Hilbert Space is a Very Big Place

qubits can be in possible states:

000..00, 000..01, 000..10, ...,
111...10, 111..11

In fact, it can be in a superposition of all of

those states at once:

127
State Vector for Two & Three Qubits

Question: How
many elements are
there in an n-qubit
state vector?

128
 Hilbert Space

Each element of the state vector

is a basis vector. The total set is
Hilbert space, and it grows
exponentially with the number of
qubits!
n.b.: Each individual qubit has two
states, or two basis vectors, but they
are multiplicative when combined --
adding a qubit adds dimensions to
the total size of our space! 129
Bell States: Our First Important Multi-Qubit States

Critical for quantum communication

Foundation for tests of entanglement
(CHSH inequality, a little later)
130
 Quantum gates

• All quantum gates are unitary.

• What’s unitary?
• said simply, reversible rotation on
Hilbert space
Simple single-qubit
gates
 X gate, Z gate

• X gate
• rotation around X axis

• Z gate
• rotation around Z axis
 Y gate

• rotation around Y axis

 Global phase
• Global phase does not affect quantum
state.

(∵ is rotation whose norm is 1 around

origin,
it does not change the distance from the
origin)
 Y gate

• rotation around Y axis

 Y gate

• rotation around Y axis

 Y gate
• rotation around Y axis
• can be created by multiplying X and Z
 Pauli gates

• X, Y, Z are called the Pauli

matrices (Pauli gates), which are
the most basic gates.
 I gate
• I gate
• Identity gate, which does not change
quantum state
 H gate
• H gate exchanges the relationship
between X axis and Z axis
 H gate
• H gate exchanges the relationship
between X axis and Z axis
Simple two-qubits
gates
 Controlled NOT
(CNOT)
• In fact, it is more correct to call controlled
X
• If the first qubit is 1, the second qubit gets
X
• the first qubit is control qubit
• the second is target qubit
 Controlled NOT
(CNOT)
• two qubit state
Circuit notation
q0 X
Z
q1

q2

q3 Z
t0 t1 t2 t3
qubits

q0 X
Z
q1

q2

q3 Z
t0 t1 t2 t3
time goes
 CNOT gate between q0 and
q1
X gate on q0 q0 is control qubit
q1 is target qubit
measurement of q0,
q0 X
Z in Z axis
q1
SWAP gate
q2 between q1 and
q2
q3 Z gate on q3
Z
t0 t1 t2 t3
Density Matrix
Representation
Pure vs. mixed states

A B

When A is part of a larger superposition, we

need to know the state of both A and B to fully
know the state of A.

If we do not have access to the state of B,

then we have lost information about A.
Pure vs. mixed states

A B

If we do not have access to the state of B,

then we have lost information about A.
If we have full information about A then we say
“A is in a pure state”.
If we do not have full information about A then
“A is in a mixed state”.
Note: we can have full information about a
superposition! is still a “pure state”.
Pure vs. mixed states

A B

Pure states (full information) are written as

state vectors, .

How do we write mixed states, where we have

lost some of the state information?

...
Representing mixed states
Representing mixed states

Density matrices let us deal with uncertainty

in the quantum state.

This uncertainty comes about because the

state we want is part of a larger superposition,
which we do not have access to.

Density matrices record the probability of

measurement outcomes.
Representing mixed states

Look again at the state

We can write the state of A using the density

matrix :
Representing mixed states

Probability of Probability of
measuring “0” measuring “1”

So diagonal elements always add up to 1.

Representing mixed states

We find the probabilities of other

measurements by changing the basis.

Let’s find the probability of measuring “+” or “-”

given this density matrix for the state.
Representing mixed states

Let’s find the probability of measuring “+” or “-”

given this density matrix for the state.

^ ^ Which operator
transforms us to the
{|+>,|->} basis?
Representing mixed states

What is this
^ ^
matrix?
Representing mixed states

So prob(0) = prob(1) = prob(+) = prob(-) = 0.5

Representing mixed states

We have seen:
The density matrix gives probabilities of
measuring the basis states.

We calculate the probabilities for different

measurements by transforming into the
basis of that measurement.

Now: how do we calculate a density matrix?

Calculating density matrices

Now: how do we calculate a density matrix?

The elements in a density matrix represent
“projection operators”:

So
Calculating density matrices

What is
^ ^ |1><1|?
Calculating density matrices
Calculating density matrices
Pure vs mixed states again

A pure state is represented as a density matrix

by a single projection operator in some basis,

Pure:
A mixed state is represented as a density matrix
by a sum of projection operators:
Mixed:
Calculating density matrices

Mixed:
Pure:
Let’s look at the state

^ ^ What is the
density
matrix for
this state?
Calculating density matrices

Mixed:
Pure:
Let’s look at the state
Calculating density matrices

Let’s look at the state

Probability of Probability of
measuring “0” measuring “1”
Diagonal elements are Off-diagonal elements
probabilities, and must are called “quantum
sum to one! coherences”
 Calculating density matrices

So is this the same

thing? Still 50/50…

This is just a classical

random variable, not a
superposition!
Representing mixed states

50/50 superposition state taken back to 0!

Representing mixed states

So prob(0) = prob(1) = prob(+) = prob(-) = 0.5!

Summary so far

A system in a mixed state is represented by a

density matrix rather than a state vector.

A pure state has a density matrix

A mixed state is a sum of pure-state density
matrices,

Still to answer: if we have a many-qubit pure

state, how can we calculate the density
matrices of the individual mixed qubit states?
 For example:
Experimental Results

D and D’ are states of

the ion, and H and V are
horizontal and vertical
polarization of the photon

A. Stute et al., Nature 482, 485 (2012)

Teleportation by the
numbers
Teleportation By the Numbers (or Symbols)

Rewrite (just
algebra!) treating
DA as a pair

and B as a solo
qubit
Teleportation By the Numbers (or Symbols)

Leaves something
BSM collapses
related to original
state to one of
D on B
these terms,
and tells us
which

Question: How do we fix it up?