Superdense coding using three qubits

Muhammad Faizan, Imama Tul Birrah Khan and Muhammad Abbas

LUMS School of Science and Engineering

March 14, 2021

Abstract

We have a situation in which Alice and Bob share a maximally entangled

GHZ state in a 3-qubit system initially. Alice can form any of the eight GHZ
states using two qubits in her possession. Sending Alice’s qubits to Bob
involves 3 bits of classical information which Bob can decode at his end by
measurement in the GHZ basis. So, 3 bits of information can be transmitted
using 2 qubits without any direct communication between parties which
would not have been possible classically. This is complementary to the 2
qubit system where Alice can prepare any of the four Bell states using 1
qubit in her possession.

1 Introduction

In quantum information theory, superdense coding is a quantum communication

protocol to transmit classical bits of information from a sender: Alice to a receiver:
Bob, under the assumption that Alice and Bob share maximally entangled state.
[1]. This protocol was first proposed by Bennett and Wiesner in 1992 and ex-
perimentally actualized in 1996 by Mattle, Weinfurter, Kwiat and Zeilinger using
entangled photon pairs [2]. By performing combination of the four quantum gates,
Alice encodes the bit and send her qubits to Bob. Classical bit is retrieved by Bob
after performing measurement in the relevant basis.

2 Protocol for tripartite system

Suppose Alice wants to send three classical bits of information (“000”,“001”,“010”,

etc.) to Bob using qubits. To do this, third party, Charles, creates maximally
entangled state of three qubits, called GHZ state and distribute qubits to Alice
and Bob. Once Alice possesses her qubits, she performs the combination of four
ˆ X̂, iŶ and Ẑ) on her qubits depending on which three-bit message
quantum gates (I,

1
she wants to send. Her entangled qubit is then sent to Bob who after applying
appropriate combination of quantum gates makes measurement in GHZ basis and
retrieves the message. Communication between Alice and Bob is unnecessary in
this case.

2.1 Preparation of GHZ state

A third party, Charles, create a maximally entangled state and distribute qubits
to Alice and Bob such that Alice has first two qubits and Bob has the third. Fig.
1 shows creation of GHZ state.

q1
H
q2

q3

Figure 1: Quantum Circuit for preparation of GHZ

Suppose input qubits are |0i ⊗ |0i ⊗ |0i. Then output state is

|000i + |111i
|ψ000 i = √
2
For different input qubits, output states are shown in Table 1.

2.2 Sharing

Now that GHZ state has been prepared, Charles will give first two qubits to Alice
and 3rd qubit to Bob because Alice will have to transform this state into any one
of the eight states given in 1 depending on the classical information she wants to
send to Bob. For example, if Alice wants to send classical bit “110”, Alice must
transform the state to |ψ110 i by applying appropriate quantum gates to her qubits.
This is only possible if Alice has two qubits in her possession.

2.3 Encoding the message

By applying combination of quantum gates to her qubits, Alice can transform

the shared entangled state to any one of the eight states. Keep in mind that
entanglment is still present.
Let’s now explain which combination of quantum gate Alice need to apply on her
qubits, depending on which 3-bit message she wants to send to Bob. Table 2 shows
the encoding rules for Alice.

2
 Sr. No Input Qubit Output State

|000i + |111i
1 |000i |ψ000 i = √
2
|001i + |110i
2 |001i |ψ001 i = √
2
|011i + |100i
3 |010i |ψ010 i = √
2
|000i − |111i
4 |100i |ψ100 i = √
2
|010i + |101i
5 |011i |ψ011 i = √
2
|001i − |110i
6 |101i |ψ101 i = √
2
|011i − |100i
7 |110i |ψ110 i = √
2
|010i − |101i
8 |111i |ψ111 i = √
2
Table 1: This table shows GHZ states corresponding to input qubits.

Note that for intended message “011”, the output state has a phase factor of eiπ
which in fact doesn’t matter.

2.4 Sending qubits

After performing appropriate gate, Alice will send her entangled qubit to Bob using
quantum network through a conventional physical medium.

2.5 Decoding

Now that Bob has all three qubits, he will measure all the qubits in GHZ basis
i.e., Bob will apply two Controlled Not (CN OT ) gates followed by Hadamard gate
(Ĥ) and then he will measure his qubit. Bob can now easily retrieve the classical
message that Alice sent. For example, if measurement results |011i, he will decode
this state to “011”.
Quantum Circuit for superdense coding is illustrated in the Fig. 2.

3
 Sr. No Intended Message Applied Gate Output state

|000i + |111i
1 “000” Iˆ ⊗ Iˆ |ψ000 i = √
2
|001i + |110i
2 “001” X̂ ⊗ X̂ |ψ001 i = √
2
|011i + |100i
3 “010” X̂ ⊗ Iˆ |ψ010 i = √
2
|000i − |111i
4 “100” Iˆ ⊗ Ẑ |ψ100 i = √
2
|010i + |101i
5 “011” Iˆ ⊗ X̂ |ψ011 i = √
2
|001i − |110i
6 “101” X̂ ⊗ iŶ |ψ101 i = √
2
|100i − |011i
7 “110” X̂ ⊗ Ẑ |ψ110 i = √
2
|010i − |101i
8 “111” Ẑ ⊗ X̂ |ψ111 i = √
2
Table 2: This table gives the rule for encoding the classical message provided that
shared entangled state is ψ000 .
Alice encodes message

U1
Sending qubits to Bob

U2
Preparation & Sharing
Bob decodes message

q1 H H b1

q2 b2

q3 b3

Figure 2: Quantum Circuit for Superdense Coding

4
2.5.1 Example

Assume that Alice and Bob share an entangled state given by:
|000i + |111i
ψ000 = √ (1)
2
First two qubits belong to Alice and third one belongs to Bob.
Suppose that Alice wants to send “011”. According to the encoding rules given in
Table. 2, Alice will apply Iˆ ⊗ X̂ to her qubits.
|ψ1 i = Iˆ ⊗ X̂ ⊗ Iˆ |ψ000 i
Iˆ |0i ⊗ X̂ |0i ⊗ Iˆ |0i + Iˆ |1i ⊗ X̂ |1i ⊗ Iˆ |1i
|ψ1 i = √
2
|010i + |101i
|ψ1 i = √ (2)
2

Alice will send her qubits to Bob via quantum network. Now Bob possesses all
three qubits. Bob will decode the message by measuring his qubits in GHZ basis.
Bob will apply two Controlled Not gates followed by Hadamard Gate as illustrated
in Fig. 2
State after first Controlled Not gate:
|011i + |101i
|ψ2 i = √ (3)
2
State after second Controlled Not gate:
|011i + |111i
|ψ3 i = √ = |+i |11i (4)
2
State after Hadamard gate:
|ψoutput i = |011i (5)

After measurement, Bob knows immediately that encoded message was “011”.
Quantum Circuit for this code is illustrated in Fig. 3.

3 Simulating the Protocol

We implemented the above explained protocol in ibmq athens by writing python

program in qiskit. Code is attached in Appendix A.

4 Results

Simulation was carried on local simulator as well as on ibmq q. Histograms of

simulators are given below:

5
 Sharing Encoding Decoding

0 H I H

0 X

000 1 2 3 output

Figure 3: Quantum Circuit for Superdense Coding

Figure 4: Probability of output state

Figure 5: Probability of output state

6
5 Conclusion

The protocol is similar to the two qubits system but not too dense. It is because of
the reason that 1.5 classical bits are transferred per qubit although total amount
of information in three qubits system is larger than two qubits system. Moreover,
it is noted that theoretically it is possible to identify any of the eight GHZ state by
means of quantum gates.In Fig. 3, GHZ state becomes disentangled at the output
ending up in three single qubits.

6 Acknowledgements

This project was carried out as a core part of PHY 414 (Quantum Computation
and Information) in the Department of Physics, LUMS, Pakistan. The authors
are thankful to Dr. M. Faryad (Chairperson), Department of Physics, LUMS,
Pakistan, for his generous support throughout the project.

References
[1] C. H. Bennett and S. J. Wiesner, “Phys. Rev. Lett.” 69, 2881 (1992).

[2] Michael A. Nielsen, Isaac L. Chaung, “Quantum Computing and Quantum

Information”, 10th Ed, Chapter 2.

7
A Appendix 1: Code
%m a t p l o t l i b i n l i n e
# Importing standard Q i s k i t l i b r a r i e s
from q i s k i t import QuantumCircuit , e x e c u t e , Aer , IBMQ
from q i s k i t . c o m p i l e r import t r a n s p i l e , a s s e m b l e
from q i s k i t . t o o l s . j u p y t e r import ∗
from q i s k i t . v i s u a l i z a t i o n import ∗
from ibm quantum widgets import ∗

# Loading your IBM Q a c c o u n t ( s )

p r o v i d e r = IBMQ. l o a d a c c o u n t ( )
####################################
qc=QuantumCircuit ( 3 , 3 )
c l a s s i c a l b i t = ” 011 ”
# GHZ P r e p a r a t i o n
qc . h ( 0 )
qc . c not ( 0 , 1 ) , qc . cno t ( 1 , 2 ) , qc . b a r r i e r ( )
# Encoding message ( e n c o d i n g r u l e s )
qc . i ( 0 ) , qc . x ( 1 ) , qc . b a r r i e r ( )
# Decoding message
qc . c not ( 1 , 2 ) , qc . cno t ( 0 , 1 ) , qc . h ( 0 ) , qc . b a r r i e r ( )
qc . swap ( 0 , 2 )
for i in range ( 3 ) :
qc . measure ( i , i )
qc . draw ( )
# use l o c a l s i m u l a t o r
from q i s k i t . v i s u a l i z a t i o n import ∗
backend = BasicAer . g e t b a c k e n d ( ’ q a s m s i m u l a t o r ’ )
s h o t s = 1024
r e s u l t s = e x e c u t e ( qc , backend=backend , s h o t s=s h o t s ) . r e s u l t ( )
answer = r e s u l t s . g e t c o u n t s ( )
print ( answer )
p l o t h i s t o g r a m ( answer )
from q i s k i t import IBMQ
from q i s k i t . p r o v i d e r s . ibmq import l e a s t b u s y

IBMQ. s a v e a c c o u n t ( ’ token ’ , o v e r w r i t e=True )

p r o v i d e r=IBMQ. l o a d a c c o u n t ( )
backend=p r o v i d e r . g e t b a c k e n d ( ’ ibmq athens ’ )
c o u n t s = e x e c u t e ( qc , backend=backend , s h o t s =1024). r e s u l t ( ) . g e t c o u n t s ( )
plot histogram ( counts )