This article has been accepted for publication in IEEE Access.

This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2024.3362955

Date of publication xxxx 00, 0000, date of current version xxxx 00, 0000.
Digital Object Identifier 10.1109/ACCESS.2023.DOI

Quantum Computing: Circuits,

Algorithms, and Applications
MUHAMMAD ALI SHAFIQUE1 , ARSLAN MUNIR2 , and IMRAN LATIF3
1
Department of Electrical and Computer Engineering, Kansas State University, Manhattan, Kansas, 66506, USA (e-mail: alishafique@ksu.edu)
2
Department of Computer Science, Kansas State University, Manhattan, Kansas, 66506, USA (e-mail: amunir@ksu.edu)
3
Brookhaven National Laboratory, U.S. Department of Energy, Upton, NY 11973-5000, USA (e-mail: ilatif@bnl.gov)
Corresponding author: Arslan Munir (e-mail: amunir@ksu.edu).

ABSTRACT Quantum computing, a transformative field that emerged from quantum mechanics and
computer science, has gained immense attention for its potential to revolutionize computation. This paper
aims to address the fundamentals of quantum computing and provide a comprehensive guide for both
novices and experts in the field of quantum computing. Beginning with the foundational principles of
quantum computing, we introduce readers to the fundamental concepts of qubits, superposition, entangle-
ment, interference, and noise. We explore quantum hardware, quantum gates, and basic quantum circuits.
This study offers insight into the current phase of quantum computing, including the noisy intermediate-
scale quantum (NISQ) era and its potential for solving real-world problems. Furthermore, we discuss the
development of quantum algorithms and their applications, with a focus on famous algorithms like Shor’s
algorithm and Grover’s algorithm. We also touch upon quantum computing’s impact on various industries,
such as cryptography, optimization, machine learning, and material science. By the end of this paper, readers
will have a solid understanding of quantum computing’s principles, applications, and the steps involved in
developing quantum circuits. Our goal is to provide a valuable resource for those eager to embark on their
quantum computing journey and for researchers looking to stay updated on this rapidly evolving field.

INDEX TERMS quantum computing, entanglement, interference, quantum circuits, quantum algorithms,
quantum applications

I. INTRODUCTION large numbers [7]. While traditional computers would require

UANTUM computing technology uses different ap- billions of years for such a task, quantum computers could
Q proaches to solve certain computational problems,
demonstrating greater efficiency compared to classical com-
potentially solve it in a short time [8], [9].
Quantum computers share some components with classical
puting systems. Recent experimental outcomes are remark- computers, such as registers, gates, and memory elements.
able, hinting at the possibility of quantum computers be- However, their underlying physical structures are funda-
coming commercially available in the near future [1]–[4]. mentally distinct and unique. Quantum computations unfold
A prominent example of quantum computing’s ability lies within quantum registers, where qubits can exist in the state
in Shor’s algorithm, renowned for its capability to factor of superposition and entanglement. These unique character-
large numbers efficiently [5]. This algorithm in 1994 marked istics make quantum computers fundamentally different from
a pivotal step in the advancement of quantum computing traditional classical computers.
by enabling the determination of prime factors of large Another distinct difference of quantum computing versus
numbers by quantum computers [6], and has also caused classical computing is computational units such as bits. Bits
consternation in cryptography field as many public key cryp- in classical computing are restricted to zero or one whereas
tography algorithms rely on the difficulty of factoring large quantum computing employ units (qubits) that are capable
numbers by classical computers. A distinct difference in of existing in states of zero, one, or any intermediate value
computational power of classical versus quantum computers [10]–[12]. This unique attribute grants quantum computers
can be illustrated with the evaluation of the time required the remarkable ability to simultaneously follow multiple
to crack encryption schemes like Rivest–Shamir–Adleman computational paths within a single calculation, which is not
(RSA) that rely on the difficulty of finding prime factors of possible by classical computers without repeated iterations.

VOLUME 4, 2023 1

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4
 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2024.3362955

Shafique et al.: Quantum Computing: Circuits, Algorithms, and Applications

A. QUANTUM COMPUTING INTRODUCTION AND achieve fault-tolerant quantum computation, substantial im-
HISTORY provements are required in quantum computers to control and
Quantum computing, in contrast to classical computing, is protect the qubits sufficiently for reliable algorithms. These
a relatively recent development. Its origins can be traced improvements can be made with hardware modifications or
back to the late 1970s when it initially appeared in science the use of error-correcting codes. Shor introduced Quantum
fiction, subsequently attracting significant attention from the Error Correction (QEC) in 1995, showing that information
media. It was in 1981 that Richard Feynman is credited with from one logical qubit can be encoded onto multiple physical
pioneering the concept of a quantum computer. He proposed qubits, protecting it from errors [22]. Shor’s work demon-
the idea that quantum computers could efficiently simulate strated the possibility of executing quantum computations
quantum systems that could avoid the exponential resource reliably with noisy quantum hardware [23].
requirements for classical computers. Classical computers Further research has revealed that noise and quantum
encounter substantial difficulties when attempting to simulate scaling relate to each other. If errors and noise are below
quantum systems. Feynman, along with visionaries like Yuri a certain threshold, it’s theoretically possible to scale up
Manin and Paul Benioff, recognized the vast potential of quantum computers to larger sizes [24], [25]. Many types
quantum computers in the realm of complicated computing of error-correcting techniques have been developed [26]–
problems. In 1985, David Deutsch formalized the concept of [28], but studies indicate that millions of physical qubits are
a quantum computer, marking a significant milestone in the needed to achieve useful quantum computers [29]. Despite
field of quantum computing. Furthermore, he distinguished this, various algorithms have claimed quantum supremacy,
between quantum simulators and programmable quantum showcasing computations on quantum devices likely surpass
devices. classical computers’ capabilities in a reasonable time frame.
In subsequent years, significant achievements were made The research works of IBM, Xanadu, and Google’s Quan-
in the field of quantum computing, revealing its potential tum AI team are prominent accomplishments in the field
to surpass classical counterparts in terms of computational of quantum computing [30]–[33]. While these achievements
efficiency. It became increasingly clear that quantum com- are significant, they have limitations in scaling up quantum
puters could offer solutions for specific computing problems computations due to noise and errors.
efficiently. Notably, Simon and Shor made remarkable con- The term NISQ stands for “Noisy Intermediate-Scale
tributions by developing algorithms that demonstrated speed Quantum.” It refers to a phase in quantum computing where
enhancements for particular problem sets, including the field quantum computers are not yet completely error-corrected
of prime factorization and cryptography. Seth Lloyd further but are large enough to perform computations beyond classi-
enriched the supremacy of quantum computers by introduc- cal computers’ capabilities. NISQ devices are characterized
ing an algorithm for simulating a wide range of quantum by the presence of errors due to noise, but they are sufficiently
systems on quantum computers. reliable for solving certain problems more efficiently than
In summary, quantum computing is getting better with classical computers [34]. This phase represents a transitional
time and has the potential to solve certain computing prob- period in the advancement of quantum computing technology
lems more efficiently than classical computers. This is shown and quantum supremacy.
by various quantum algorithms, which highlight how power- The remainder of this paper is organized in the following
ful quantum computing can be. manner. Section II provides the fundamental concepts of
qubits, superposition, entanglement, interference, and noise.
B. NOISY INTERMEDIATE-SCALE QUANTUM (NISQ) Section III explores the building blocks of quantum circuits
From the beginning, there has been a doubt whether a quan- such as quantum gates and measurement. In Section IV, vari-
tum computer could surpass the capabilities of a classical ous types of quantum circuits are explained with numerical
computer. Many of these doubts originate from concerns examples. Section V discusses quantum algorithms which
about the complexity of quantum computer design and and include famous Shor’s algorithm and Grover’s algorithm.
difficulty of controlling quantum computation devices [13]– Section VI elaborates the utilization of quantum computing
[15]. These concerns are primarily related to the concept in machine learning. Popular quantum simulators and their
of decoherence, where quantum systems interact with their features are explored in Section VII. Quantum hardware and
environment and lose their quantum properties (superposi- its deployment requirements are discussed in Section VIII.
tion, entanglement, and, interference) over time affecting the Section IX discusses quantum computing applications for
outcome of quantum circuits. A controlled quantum environ- various fields, such as cryptography, optimization, machine
ment has led to debates about achieving reliable quantum learning, and finance. Lastly, Section X concludes the ar-
computers [16]. It also suggests that quantum computers can ticle by summarizing the fundamental concepts, algorithms
outperform classical if certain conditions are met [17]. and applications of quantum computing. It also motivates
While early noisy quantum computers have been used the researchers to apply this rapidly evolving technology in
to implement algorithms such as Shor’s, Grover’s, and different fields.
Deutsch–Jozsa’s, the prevailing high error rates and noise
prevent the scaling of these algorithms [18]–[21]. In order to II. QUANTUM COMPUTING FUNDAMENTALS

2 VOLUME 4, 2023

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4
 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2024.3362955

Shafique et al.: Quantum Computing: Circuits, Algorithms, and Applications

A. QUBITS A single qubit is also called a two-level quantum system

In classical computing, a bit is analogous to a binary light because it is a linear combination of two state basis, 0 and
switch, capable of assuming only two discrete states: 0 or 1. Below is the common form of a single qubit in bra-ket
1, without any intermediary values. In contrast, for quan- notation.
tum computing, a quantum bit (qubit) operates more like
" # " # " #
1 0 v0
a dimmer switch. It possesses not just the 0 and 1 states |v⟩ = v0 |0⟩ + v1 |1⟩ = v0 + v1 = (5)
but also the ability to exist in an intermediate state, which 0 1 v1
is a linear combination of the 0 and 1 states, weighted by Here vo and v1 are complex coefficients to measure prob-
specific coefficients. These coefficients are used to calculate ability amplitudes. The probability and the phase of each
the probability of measuring either the 0 or 1 state when computational state basis for a qubit can be computed as
measured. follows:
For state basis |0⟩ with complex coefficient vo = x + i ∗ y
1) Bra-ket notation
Qubit is a quantum computing particle that has a wave-like probability amplitude = |vo | (6)
nature with wavefunction ψ(x) that satisfies the Schrödinger p p
|vo | = (x + i · y) ∗ (x − i · y) = x2 + y 2 (7)
equation. Theoretically, this wavefunction exists in an in-
finite dimensional Hilbert dual space [35]. Therefore, the probability = |vo |2 (8)
state vector representing this wavefunction in Hilbert space y
requires an infinite dimensional vector notation. This infinite phase(rad) = tan−1 (9)
x
dimensional vector state of the qubit in Hilbert dual space
is shown using Dirac’s bra-ket notation, which was created phase (degree) = phase (rad) ∗ (180/π) (10)
by Paul Dirac in 1939 [36]. However, it can also be a finite-
The probability amplitude is used to calculate the proba-
dimensional vector having two states, on/off or spin-up/spin-
bility of each state basis of the qubit which helps in the
down, which can be shown in two-dimensional Hilbert space.
measurement of the qubit state. Similarly, phase is used for
In this notation, two-dimensional state vectors |1⟩ (read ket
quantifying interference. The concepts of measurement and
one) and |0⟩ (read ket zero) are used for qubit.
interference are explained in the following sections of the
" # paper. If the complex coefficients are normalized, then they
1
|0⟩ = 1|0⟩ + 0|1⟩ → (1) represent the probability of the qubit for 0 and 1 state
0
" # |vo |2 + |v1 |2 = 1 (11)
0
|1⟩ = 0|0⟩ + 1|1⟩ → (2) This is known as the normalization constraint since all
1 two-level systems must obey this quality to function as a
In the equation (1), ket zero shows that the qubit is at an qubit.
off or spin-down state. Here, the first element represents the For two or multiple qubits, the tensor product (or Kro-
probability amplitude of off or spin down, and the second necker product) is used to compute the resultant states of
element shows the probability amplitude of on or spin up. the quantum system. The tensor product is denoted by the
Probability amplitude can be a complex value and it is used symbol ⊗. Let us consider two qubits |a⟩ and |b⟩ as
to compute the probabilities of vector states. Additionally, in
" # " #
a0 b0
Dirac’s notation, the bra is a complex conjugate transpose of |a⟩ = and |b⟩ = (12)
a1 b1
a ket. For example, ⟨ϕ| (read bra of ϕ) is a complex conjugate
transpose vector of ket ψ. The inner product of these two The tensor product of the two qubits is
vectors ⟨ϕ|ψ⟩ is a scalar value [37]. The symbol “|⟩” denotes
a column vector, and is known as a “ket”. The “bra” (⟨|) form |x⟩ = |a⟩ ⊗ |b⟩ = |ab⟩ (13)
is a row vector and it is shown below:
 " #
h i b0
a0 ∗
   
⟨0| = 1⟨0| + 0⟨1| → 1 0 (3) a0 b0 x0
b1  


  a0 b1  x1 
 
|x⟩ =  = =  (14)

h i   a b  x2 
1 0
" #
⟨1| = 0⟨0| + 1⟨1| → 0 1 (4)
 
b0 
 
a1 ∗ a1 b1 x3

The ket notation is widely used in quantum computing as b1
bra-ket representation
√ of the qubit.
√ The following two states
(|0⟩ + |1⟩)/ 2 and (|0⟩ − |1⟩)/ 2 are also commonly used |x⟩ = a0 b0 |00⟩ + a0 b1 |01⟩ + a1 b0 |10⟩ + a1 b1 |11⟩ (15)
in quantum calculations and these are sometimes written as
|+⟩ and |−⟩, respectively. |x⟩ = x0 |00⟩ + x1 |01⟩ + x2 |10⟩ + x3 |11⟩ (16)
VOLUME 4, 2023 3

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4
 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2024.3362955

Shafique et al.: Quantum Computing: Circuits, Algorithms, and Applications

and the normalization constraint rule for the two qubits Conversely, in quantum computing, a single quantum bit
will be the same as follows: (qubit) can exist in the state of 0, 1, or any linear combination
of these states as shown in Figure 2. This phenomenon is
|a0 b0 |2 + |a0 b1 |2 + |a1 b0 |2 + |a1 b1 |2 = 1 (17) called superposition which enables qubits to exist in a com-
" # bination of the states. Upon measurement, the superposition
c0 collapses, and the final outcome is determined depending
Similarly for 3-qubits, if |c⟩ = then the tensor
c1 on the probability distribution of the qubit states. Quantum
product of the three qubits is superposition is the ability of a qubit to be in multiple states
simultaneously until it is measured.
|y⟩ = |ab⟩ ⊗ |c⟩ = |abc⟩ (18)
C. QUANTUM ENTANGLEMENT
 " # In classical computers, the state of a bit can vary indepen-
c0 dently, that is, the state of a bit is not influenced by the state of
a0 b0 ∗
c1  another bit. However, in quantum computing, the probability


of a qubit state can be affected by the change of another qubit
     
  a0 b o c0 y0
state probability. This phenomenon is called entanglement
 
" # 
c0  a0 bo c1  y1 
   
a0 b1 ∗ [38]. In quantum circuits, entanglement is created through

    

 c1 
 a0 b1 c0  y2 
   
quantum gates by performing specific operations on the
 a b c  y 
qubits that result in inseparable states of qubits as shown

 0 1 1  3
|y⟩ =  = =  (19)
 " #
 c0  a1 bo c0  y4  in equations (21), (22), (23), and (24). Regardless of the
a1 b0 ∗
physical distance between the entangled qubits, a change
    
 a1 bo c1  y5 
c1 
    
in one qubit state probability can change the probability

 a1 b1 c0  y6 
     
distribution of all qubits in the entangled quantum system

 " #
a b ∗ c0 
  a1 b1 c1 y7 [39].
 1 1
c1 

 Quantum entanglement is a phenomenon that occurs when
two or more particles become correlated in such a way that
the state of one qubit is dependent on the state of the other
The same method will be used to combine n qubits, and qubit, regardless of the distance between them. If the state
normalization constraint rules for n-qubits will be given as in of one qubit changes in the entangled system, then the states
equation (20). X of all other qubits will be affected. There are specific states
|vi |2 = 1 (20) in 2-qubit systems, which are called Bell’s states or EPR
(Einstein–Podolsky–Rosen) pairs, which exhibit entangled
If we have n qubits, we will need to keep track of 2n complex properties and cannot be written in separable states as given
probability amplitudes. As we can see, these vectors grow below:
exponentially with the number of qubits. This is the reason
quantum computers with large numbers of qubits are so
1
difficult to simulate in classical computers. A modern laptop |ϕ⟩ = √ (|0⟩|0⟩ + |1⟩|1⟩) (21)
can easily simulate a general quantum state of around 20 2
qubits, but simulating 100 qubits is too difficult even for the
largest supercomputers.
′ 1
|ϕ ⟩ = √ (|0⟩|0⟩ − |1⟩|1⟩) (22)
2
2) Bloch sphere notation
′′ 1
The Bloch sphere is a mathematical representation of a given |ϕ ⟩ = √ (|0⟩|1⟩ + |1⟩|0⟩) (23)
quantum state of a qubit, with which researchers can pinpoint 2
and manipulate various such states within the sphere to their
advantage. Three qubits |1⟩, |−⟩ and, |γ⟩ are shown in Bloch ′′′ 1
|ϕ ⟩ = √ (|0⟩|1⟩ − |1⟩|0⟩) (24)
sphere representation in Figure 1. 2
In the 2-qubit system, as shown in Figure 3a, each of the
B. QUANTUM SUPERPOSITION qubits is in a superposition state but qubits are not entangled.
In classical computing, a bit possesses a binary nature, ex- Therefore, the probabilities of all superposed qubit states are
clusively adopting either a state of 1 or 0. Correspondingly, independent of each other. When these qubits are entangled
in a 2-bit classical system, only one state can exist at a given as shown in Figure 3b, then the change in the probability of
time among four distinct states that is 00, 01, 10, and 11. one qubit affects the probabilities of the entangled qubits.
This conceptual framework can be extended to n-bit classical The blue color in Figure 3b shows that the two qubits are
systems with 2n states but only one state exists at a given not independent particles. They are entangled and their states
time representing the state of the classical system. are dependent on each other. This entanglement results in the
4 VOLUME 4, 2023

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4
 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2024.3362955

Shafique et al.: Quantum Computing: Circuits, Algorithms, and Applications

" # " # " #

0 1 0.8137 + 0.0i
(a) |1⟩ = (b) |−⟩ = √1 (c) |γ⟩ =
1 2 −1 0.0 + 0.5812i

FIGURE 1: Bloch sphere representation of three different qubits.

(a) (b) (c) FIGURE 4: Wave-like nature of the qubit

FIGURE 2: Representation of qubit with the state of 0 in (a),
1 in (b), and superposed states in (c) When we have multiple qubits, their wavefunctions are
added together to give an overall wavefunction describing the
change of probability distribution of the state of the entangled resultant states of a quantum system. This adding process of
quantum system, even if the entangled qubits are far away wavefunctions is called interference. It is a fundamental phe-
from each other. nomenon that arises from the wave-like nature of quantum
particles, such as electrons or photons and it distinguishes
quantum systems from classical systems.
In quantum computing, when two quantum wavefunctions
overlap, they can interfere with each other constructively or
destructively. This results in a change in the resultant wave-
function of the quantum system that affects the probability
distribution of its quantum states as shown in Figure 5.

(a) (b)

FIGURE 3: Two qubits in non-entangled (a) and entangled

quantum states (b)

FIGURE 5: Quantum interference in two quantum wavefunc-

D. QUANTUM INTERFERENCE tions
Qubit is represented with bra-ket notation or Bloch sphere but
this is just a mathematical representation of the qubit state. In Interference can also be a challenge in quantum computing
reality, the qubit has a wave-like nature that is described by a due to the phenomenon of decoherence. Decoherence is the
quantum wavefunction satisfying the Schrödinger equation loss of quantum coherence, which is the property of a quan-
as shown in Figure 4. A wavefunction is a mathematical tum state to maintain its superposition and entanglement, due
description of the quantum state that consists of complex to the interactions with environment and thermalization. This
probability amplitudes, and the corresponding probabilities loss of coherence leads to a breakdown of interference effects
of quantum system states. and making quantum computation error-prone. Quantum er-
VOLUME 4, 2023 5

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4
 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2024.3362955

Shafique et al.: Quantum Computing: Circuits, Algorithms, and Applications

ror correction techniques are used to mitigate the impact of distribution of states within a quantum circuit. Quantum
the external environment and preserve the delicate quantum algorithms are constructed by applying a sequence of gates
interference necessary for quantum computation. to qubits in a specific order. Ultimately, the measurement is
Overall, interference is a foundational concept in quantum executed at the end of the quantum circuit, collapsing the
computing, allowing quantum systems to perform computa- superposition states and revealing the quantum circuit’s state
tions by updating the probability distributions of the quantum based on the probability distribution of the states.
states. This concept solves certain problems in ways that are
not achievable using classical computing methods.

E. QUANTUM NOISE
Quantum noise refers to the uncertainty and fluctuations that
arise in quantum systems due to the probabilistic nature of
quantum mechanics. It is a challenge in quantum systems
even at low temperatures.
In classical systems, noise is often associated with ran-
dom variations in signals or disturbances caused by external
factors. When a quantum system is in a superposition state,
FIGURE 6: Quantum circuit example
its outcome upon measurement is not deterministic but is
determined by the probability distribution of the quantum
states. Noise and error can affect the outcome due to the
B. QUANTUM GATES
quantum system’s interaction with the external environment.
Quantum gates are basic building blocks in quantum comput-
It can lead to loss of quantum properties (superposition,
ing that manipulate the state of the qubits in quantum circuits.
entanglement, and, interference) over time affecting the out-
They are analogous to classical logic gates in classical com-
come of quantum circuits.
puting but leverage the principles of quantum mechanics for
Quantum noise has several manifestations in quantum
processing quantum information. Quantum gates are summa-
systems, and it can impact various aspects of quantum com-
rized in Table 1 with symbol and matrix representation.
puting. Some common examples of quantum noise include:
• Measurement Noise: When measuring a quantum sys- 1) Identity Gate
tem, the act of measurement can cause a quantum sys- The identity gate is a single-qubit gate. It does not modify
tem to lose the quantum superposition and collapse the the state of a qubit; therefore, it is represented by an identity
quantum state into one of its states, introducing uncer- matrix as shown in equation (25).
tainty in the outcome due to the probabilistic nature of " #
the measurement process. 1 0
I= → (25)
• Decoherence: Interactions with the environment can 0 1
cause quantum systems to lose quantum superposition,
entanglement, and interference, affecting the perfor- The identity gate is helpful during mathematical computa-
mance of quantum algorithms. tion. This gate is used to compute the resultant gate matrix
in multi-qubit circuits. The identity matrix is an involutory
Quantum noise poses a significant challenge for quantum
matrix, meaning that the identity matrix is equal to its inverse
computing. To address this challenge, researchers have been
matrix as shown in equation (26).
working on quantum error correction techniques, which are
essential for preserving the quantum states against the detri- I = I −1 I2 = I (26)
mental effects of measurement noise and decoherence.
2) Single-Qubit Pauli Gates
III. BUILDING BLOCKS OF QUANTUM CIRCUITS The Pauli gates (X, Y, and Z) are based on Pauli matrices
A. HOW YOU PERFORM COMPUTATION WITH QUBITS (σx , σy , σz ), which are useful to modify the state or phase of
Practically all classical computers work in a similar way. the qubit. Pauli gates are single-qubit gates and they rotate
They use a group of bits to store information in a binary form the state of the qubit around the x, y, and z axes of the Bloch
within memory. The state of these bits can be altered using sphere by π radians.
logical gates like AND, OR, NOT, and NAND gates. • Pauli-X gate is the quantum equivalent of the NOT gate
Quantum computing utilizes a model similar to classical for classical computers. It maps |0⟩ to |1⟩ and |1⟩ to |0⟩.
computing known as the gate model or circuit model as It is also called a qubit-flip gate.
shown in Figure 6. Within this framework, a set of qubits • Pauli-Y maps |0⟩ to i|1⟩ and |1⟩ to −i|0⟩.
is used and their states can be superposed and entangled • Pauli-Z maps |1⟩ to −|1⟩ and leaves the state |0⟩ un-
with each other. Multiple gates are available for conducting changed. Due to this nature, Pauli Z is also called a
operations on these qubits, thereby altering the probability phase-flip gate.
6 VOLUME 4, 2023

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4
 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2024.3362955

Shafique et al.: Quantum Computing: Circuits, Algorithms, and Applications

These matrices are usually represented as  

1 0 0 0
" # 0 1
0 1 0 0 
→ CU =  (36)
 
X= (27) 
1 0 0 0 u00 u01 
0 0 u10 u11
" #
0 −i where U is one of the Pauli matrices that is X, Y, and Z. Based
Y = → (28) on the respective Pauli matrices, we can have "controlled-
i 0
X", "controlled-Y", or "controlled-Z" gates. The shortened
" # symbols of these controlled gates are CX, CY, and CZ
1 0
Z= → (29) respectively.
0 −1 Controlled-Y gate (CY) performs the Y operation on the
second qubit only when the first qubit is |1⟩, and otherwise
The Pauli matrices are anti-commute as shown in equa-
leaves it unchanged.
tion (30). Two matrices A and B are considered to be anti-  
commute if AB = -BA. 1 0 0 0
0 1 0 0 
ZX = −XZ = iY (30) CY = 

 →

(37)
0 0 0 −i
The Pauli matrices are traceless matrices meaning the sum 0 0 i 0
of the eigenvalues of Pauli matrices is zero.
" # Controlled-Z gate (CZ) performs the Z operation (phase
−1 flip) on the second qubit only when the first qubit is |1⟩, and
eig(X) = eig(Y ) = eig(Z) = (31) otherwise leaves it unchanged.
1
 
The Pauli matrices are also involutory, meaning that Pauli 1 0 0 0
0 1 0 0 
matrices are equal to their inverse matrices as shown in CZ =   → (38)
 
equation (32), and the square of the Pauli matrices is the 0 0 1 0 
identity matrix as shown in equation (33),. 0 0 0 −1
X = X −1 Y = Y −1 Z = Z −1 (32) The Controlled gates are Hermitian matrices as shown in
equation (39). A square matrix (complex or real) is said to
X2 = Y 2 = Z2 = I2 (33) be a Hermitian matrix if it is equal to its own conjugate
transpose matrix.
3) Two-Qubit Controlled Gates
Controlled gates are two-qubit gates, where the first qubit CX = CX T CY = CY T CZ = CZ T (39)
acts as a control and the second qubit is a target qubit. The Controlled gates are also involutory, meaning that
The controlled-NOT gate (CNOT) also known as controlled- Controlled gates matrices are equal to their inverse matrices
X gate (CX), operating on two qubits, performs the NOT as shown in equation (40), and the square of a Controlled
operation on the second qubit (target qubit) only when the gates matrices is the identity matrix as shown in equa-
first qubit (control qubit) is |1⟩ otherwise leaves the second tion (41).
qubit unchanged. It is represented by the matrix CNOT.
CX = CX −1 CY = CY −1 CZ = CZ −1 (40)
  CX 2 = CY 2 = CZ 2 = I 2 (41)
1 0 0 0
0 1 0 0
CN OT = 
 
 → (34) 4) Single-Qubit Phase Shift Gates
0 0 0 1 The phase shift gates are quantum gates that introduce a
0 0 1 0 phase shift to the quantum state of a qubit. These are one-
qubit gates and these are used to alter the phase of the qubit’s
Controlled gates can be generalized using the universal ma- state without changing its probability amplitudes. They map
trix U which is a single-qubit matrix. the states |0⟩ → |0⟩ and |1⟩ → eiφ |1⟩. The probability of
" # measuring a |0⟩ or |1⟩ does not change after applying this
u00 u01 gate. Generally, the phase shift gate is generally represented
U= (35) by the matrix P:
u10 u11
" #
The controlled-U gate acting on two qubit can be defined 1 0
P = (42)
as CU. 0 eiφ
VOLUME 4, 2023 7

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4
 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2024.3362955

Shafique et al.: Quantum Computing: Circuits, Algorithms, and Applications

where φ is the phase shift with the period 2π. Some common The SWAP matrix is an involutory matrix as shown in
examples of phase shift gates are, equation (53).
π
• T gate where φ = 4
SW AP = SW AP −1 SW AP 2 = I (53)
" #
π 1 0
T = P( ) = π → (43) 7) Three-Qubit Toffoli (CCNOT) Gate
4 0 ei 4 The Toffoli gate is named after Tommaso Toffoli and is also
• S gate, though S notation is sometimes used for SWAP known as CCNOT gate (Controlled-Controlled-NOT gate) or
gate where φ = π2 Deutsch gate D(π/2). The Toffoli gate is like a CNOT gate
" # with two control qubits and one target qubit. The target qubit
π 1 0 will be inverted if the first and second qubits are in |1⟩ state.
S = P( ) = π → (44) It is represented by the matrix CCNOT as given below
2 0 ei 2
• The Pauli-Z gate where φ = π  
1 0 0 0 0 0 0 0
0 1 0 0 0 0 0 0
" #  
1 0 
0

Z = P (π) = → (45) 0 1 0 0 0 0 0
0 eiπ 
0

0 0 1 0 0 0 0
CCN OT =  →
 
0 0 0 0 1 0 0 0
 
5) Single-Qubit Hadamard Gate 0
 0 0 0 0 1 0 0
The Hadamard or Walsh-Hadamard gate, named after 0

0 0 0 0 0 0 1

Jacques Hadamard and Joseph L. Walsh, is applied on a
0 0 0 0 0 0 1 0
single qubit. It creates superposition states with an equal (54)
probability distribution for a given qubit. It maps the state The CCNOT matrix is an involutory matrix as shown in
of the qubit as follows: equation (55).
|0⟩ + |1⟩
|0⟩ → √ (46) CCN OT = CCN OT −1 CCN OT 2 = I (55)
2
|0⟩ − |1⟩ C. MEASUREMENT
|1⟩ → √ (47)
2 In quantum computing, measurement is a fundamental oper-
The Hadamard gate is represented by a matrix H: ation that provides a way to extract information from a quan-
" # tum system. When a quantum measurement is performed
1 1 1 on a qubit, it yields a classical result. This measurement
H=√ → (48)
2 1 −1 outcome collapses the superposition states of the qubit and
provides either 0 or 1 based on the probability distribution
The Hadamard matrix is a traceless and involutory matrix of the qubit states. Measurement provides information about
as shown in equation (49) and equation (50) respectively. the quantum state of the qubit at the time of measurement.
" √ #
− 2 Measurements convert multiple (superposition) probabilistic
eig(H) = √ (49) states to one absolute state and collapse other states and
2
superpositions. This is known as the measurement problem
of quantum mechanics.
H = H −1 H2 = I (50)
IV. BASIC QUANTUM CIRCUITS
6) Two-Qubit Swap Gate A quantum circuit is a series of qubits and gates. Qubits
A swap gate performs a swap operation between two qubits. can be in superposed or entangled states. Gates are used to
It is a fundamental quantum gate used in quantum computing change the state of qubits. Several gates perform different
with the primary purpose of exchanging the states of two operations that are summarized in Table 1. Quantum gates are
qubits. represented in matrix form, and the qubit’s states are denoted
SW AP |qubit1, qubit2⟩ = |qubit2, qubit1⟩ (51) in vector notation. The overall state of the quantum system
can be calculated using the matrix product between the gate’s
It is represented by the matrix: matrices and qubit vectors. Common examples of quantum
  circuits are summarized in Figure 7.
1 0 0 0
0 0 1 0
SW AP = 
 
→ (52) A. SINGLE-QUBIT GATE ON SINGLE QUBIT

0 1 0 0 In a single-qubit gate operated on a single qubit, the order of
0 0 0 1 the gate matrix will be 2×2 which is equal to the total number
8 VOLUME 4, 2023

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4
 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2024.3362955

Shafique et al.: Quantum Computing: Circuits, Algorithms, and Applications

TABLE 1: Common Quantum Logic Gates

Common Quantum Logic Gates

Operator Gate Number of Qubit Gate Symbol Matrix
" #
1 0
Identity gate One-qubit
0 1
" #
0 1
Pauli-X (X) One-qubit
1 0
" #
0 −i
Pauli-Y (Y) One-qubit
i 0
" # " #
1 0 1 0
Pauli-Z (Z) One-qubit =
0 −1 0 eiπ
 
1 0 0 0
0 1 0 0
Controlled-X (CNOT,CX) Two-qubit
 
0 0 0 1
 

0 0 1 0 
1 0 0 0
0 1 0 0
Controlled-Y (CY) Two-qubit
 
0 0 0 −i
 
0 0 i 0
 
1 0 0 0
0 1 0 0 
Controlled-Z (CZ) Two-qubit
 
0 0 1 0 
 
0 0 0 −1
" #
1 0
Phase Gate-S One-qubit π
0 ei 2
" #
1 0
Phase Gate-T (π/8) One-qubit π
0 ei 4
" #
1 1
Hadamard (H) One-qubit √1
2 1 −1
 
1 0 0 0
0 0 1 0
SWAP Gate Two-qubit
 
0 1 0 0
 
0 0 0 1
1 0 0 0 0 0 0 0
 
0 1 0 0 0 0 0 0
 
0 0 1 0 0 0 0 0
 
0 0 0 1 0 0 0 0
Toffoli Gate (CCNOT, CCX, Three-qubit
 
0 0 0 0 1 0 0 0
 
TOFF)  
0 0 0 0 0 1 0 0
 
0 0 0 0 0 0 0 1
0 0 0 0 0 0 1 0

of states for the 1-qubit quantum system, that is, two. The to the total number of states for a 2-qubit quantum system.
output state of the quantum system can be calculated using Similarly, for a multiple qubits gate operated on three qubits,
the matrix product of the gate matrix and the vector notation the order of the gate matrix would be 8×8 which is equal to
of the qubit. NOT gate and Hadamard gate are examples of the total number of states for the 3-qubit quantum system and
single gate applied on a single qubit as shown in Figures 7a so on. The resultant input quantum state for multiple qubits
and 7b, respectively. can be calculated using the Kronecker product (or tensor
product) of all qubits of the quantum system as shown in
equation (58). For two or multiple qubits, the tensor product
B. MULTIPLE-QUBIT GATE ON MULTIPLE QUBITS
is used to find the resultant quantum states. The tensor
In a multiple qubits gate operated on multiple qubits, the product is denoted by the symbol ⊗.
order of the gate matrix depends on the number of qubits Let us consider qubits |a⟩ and |b⟩:
it operates. For a multiple qubits gate applied on two qubits,
the order of the gate matrix would be 4×4 which is equal
VOLUME 4, 2023 9

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4
 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2024.3362955

Shafique et al.: Quantum Computing: Circuits, Algorithms, and Applications

" # " # system is computed using the second method as explained

ao bo above. If the number of single-qubit gates are less than the
|a⟩ = and |b⟩ = (56)
a1 b1 total number of qubits, then identity gate can be used for the
computation of the resultant multiple-qubit gate.
The resultant input quantum state for multiple qubits can
be calculated using the tensor product of all qubits of the D. ENTANGLED CIRCUITS
quantum system. In classical systems, only one state exists at a time but in
|α⟩ = |a⟩ ⊗ |b⟩ (57) quantum computing, all states can exist simultaneously. This
combination of the states is called superposition in quantum
" #
 systems. These superposed states continue if there is no
bo
ao ∗ external influence. However, when the superposed states are
 
ao bo
b1  


  ao b1  measured externally, the superposition collapses and the final
|α⟩ = |ab⟩ =  = (58)

  a b  output state is recorded based on the probability distribution
1 o
" #

bo 
 of the qubit states. Superposed states of the qubit can also be
a1 ∗ a1 b1

b1 entangled states or not. It can be understood with an example.
Let |v⟩ represents a 2-qubit quantum system as

|α⟩ = ao bo |00⟩ + ao b1 |01⟩ + a1 bo |10⟩ + a1 b1 |11⟩ (59) |v⟩ = a0 b0 |00⟩ + a0 b1 |01⟩ + a1 b0 |10⟩ + a1 b1 |11⟩ (67)

|v⟩ = |a⟩ ⊗ |b⟩ = v0 |00⟩ + v1 |01⟩ + v2 |10⟩ + v3 |11⟩ (68)

|α⟩ = αo |00⟩ + α1 |01⟩ + α2 |10⟩ + α3 |11⟩ (60)
where |v0 |2 ,|v1 |2 ,|v2 |2 , and |v3 |2 are the probabilities of |v⟩
The output state of the quantum system can be calculated qubit states.
using the matrix product of the gate matrix and vector nota- If |v⟩ can be written as
tion of the input quantum state of the gate. SWAP gate and
CNOT gate are examples of multiple-qubit gates applied on |v⟩ = (a0 |0⟩ + a1 |1⟩) × (b0 |0⟩ + b1 |1⟩) = |a⟩ ⊗ |b⟩ (69)
multiple qubits shown in Figures 7c and 7d respectively.
then |v⟩ is not entangled because |v⟩ can be written as sep-
arable states of qubits |a⟩ and |b⟩. The probabilities of qubit
C. LOWER ORDER QUBIT GATE ON HIGHER # QUBITS
|a⟩ and |b⟩ states are independent of each other. However,
In this case, the order of the gate matrix is less than the total if |v⟩ cannot be written into separable states then these are
states of the quantum system. In such a scenario, two meth- entangled states. Figure 7f shows an example of an entangled
ods are possible. In the first method, the lower order qubit quantum circuit with calculations and probabilities amplitude
gates g0 and g1 are applied to the qubits individually. The plot.
output state |C⟩ of the quantum system is computed using the Non-Entangled System Example:
Kronecker product of the output vectors |c0⟩ and |c1⟩ which Let us take an example of a quantum system that is not entan-
are calculated from each lower order qubit gate as shown in gled and see the effect of measurement on the probabilities of
equation (63). 2-qubit quantum states:
|c0⟩ = g0 × |a⟩ (61)
1
|c1⟩ = g1 × |b⟩ (62) |ϕ1⟩ = √ (|0a ⟩|0b ⟩ + |0a ⟩|1b ⟩) (70)
2
|C⟩ = |c0⟩ ⊗ |c1⟩ (63) the probability of |a⟩ as |0a ⟩ before measuring |b⟩ is
The second method is to compute the resultant input state |v⟩ 1 1
prob(|0a ⟩) = ( √ )2 + ( √ )2 = 1 (71)
of the quantum system using the Kronecker product of the 2 2
qubits, that is: Suppose if we measure the state of |b⟩ as |1b ⟩ then the
|v⟩ = |a⟩ ⊗ |b⟩ (64) superposed states of |ϕ1⟩ will be collapsed and we end up
G = g0 ⊗ g1 (65) with

|C⟩ = G × |v⟩ (66) |ϕ1⟩ = |0a ⟩|1b ⟩ (72)

The resultant quantum gate G can also be calculated using Now the probability of |a⟩ as |0a ⟩ after measuring |b⟩ is
the Kronecker product of lower order qubit gates as shown prob(|0a ⟩) = (1)2 = 1 (73)
in equation (65). The output quantum state can be found
using the matrix product of the input quantum state and the The probability of |a⟩ states have not been affected by the
corresponding gate matrix as shown in equation (66). change in |b⟩ state therefore it is a non-entangled quantum
In Figure 7e, 1-qubit Hadamard and NOT gates are applied system.
on a 2-qubit quantum system. The output state of the quantum Entangled System Example:
10 VOLUME 4, 2023

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4
 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2024.3362955

Shafique et al.: Quantum Computing: Circuits, Algorithms, and Applications

Now let us take an example of a quantum system that is en- This algorithm has potential applications in cryptography
tangled and see the effect of measurement on the probabilities and demonstrates the advantages of quantum computing for
of 2-qubit quantum states: certain problems.
1
|ϕ2⟩ = √ (|0a ⟩|0b ⟩ + |1a ⟩|1b ⟩) (74) C. SIMON’S ALGORITHM
2
Simon’s algorithm is a quantum algorithm for obtaining
The probability of |a⟩ as |0a ⟩ before measuring |b⟩ is the period of a vectorial Boolean function with polyno-
1 mial time complexity. The algorithm achieves exponential
prob(|0a ⟩) = ( √ )2 = 0.5 (75) speedup over classical algorithms. It has applications in
2
quantum cryptanalysis and cryptography. Simon’s algorithm
Suppose if we measure the state of |b⟩ as |1b ⟩ then the has been utilized to study the autocorrelation spectrum and
superposed states of |ϕ2⟩ will be collapsed and we end up Walsh spectrum of Boolean functions [45]. It has also been
with applied in the design of a lightweight encryption algorithm
|ϕ2⟩ = |1a ⟩|1b ⟩ (76) called SIMON-GCM for IoT security, which combines the
SIMON cipher block and Galois/Counter Mode (GCM) [46].
Now the probability of |a⟩ as |0a ⟩ after measuring |b⟩ is Additionally, this algorithm has been analyzed for different
prob(|0a ⟩) = 0 (77) use cases in cryptanalysis [47].

Here, the probability of |a⟩ state has been changed by the D. SHOR’S ALGORITHM
change in |b⟩ state, therefore, it is an entangled quantum One of the most compelling quantum algorithm is Shor’s
system. algorithm [5], which is a quantum algorithm for finding the
prime factors of an integer. It was developed in 1994 by the
V. QUANTUM ALGORITHMS American mathematician Peter Shor [48]. It has the potential
A. DEUTSCH-JOZSA ALGORITHM to break widely used encryption schemes, such as RSA,
The Deutsch-Jozsa algorithm is a quantum algorithm that which rely on the difficulty of factoring large numbers. It
solves the Deutsch-Jozsa problem, which involves deter- leverages the quantum properties of superposition and entan-
mining whether a given Boolean function is balanced or glement to perform the factorization process exponentially
constant. The algorithm can solve this problem with just one faster than the best-known classical algorithms. In the integer
step, providing a significant speedup compared to classical factorization problem, given an integer N = p × q for some
algorithms. It was proposed by David Deutsch and Richard prime numbers p and q, the main goal is to find the prime fac-
Jozsa in 1992 and is one of the early examples of a quantum tors p and q. The traditional
√ trial division method has a time
algorithm that provides a significant speedup over classical complexity of about O( N ) and the best classical algorithm
algorithms. The algorithm generalizes the Deutsch algorithm is the general number field sieve (GNFS), which has a sub-
to handle multiple degrees of freedom and can be derived exponential time complexity. For GNFS, the time complexity
from the quantum Fourier transform algorithm [40]. Ad- is roughly O(exp((64/9)1/3 ×(logN )1/3 ×(log∗logN )2/3 ))
ditionally, the algorithm has been used as an educational [49]. Shor’s quantum algorithm solves this problem substan-
experiment to demonstrate qubit fundamental concepts and tially faster, in time O(logN )3 ). It’s important to note that the
algorithmic challenges in quantum science and technology implementation of Shor’s algorithm becomes more complex
[41]. Experimental implementations of the Deutsch-Jozsa for large numbers. In practice, Shor’s algorithm demonstrates
algorithm have also been performed on IBM’s quantum the potential of quantum computing to break widely used
computer, showcasing its efficiency compared to classical encryption methods, emphasizing the need for post-quantum
techniques [42]. cryptography techniques.

B. BERNSTEIN-VAZIRANI ALGORITHM E. GROVER’S ALGORITHM

The Bernstein-Vazirani algorithm is a quantum algorithm Grover’s algorithm is a quantum algorithm designed to find a
designed to find a hidden binary string in a function. It uses specific value within unsorted (or unstructured) databases or
quantum principles to extract information about the hidden lists exponentially faster than classical algorithms. It provides
string and determine it with fewer queries than classical al- a quadratic speedup,√ allowing it to search the target item in
gorithms, even in the presence of noise and imperfect equip- approximately O( N ) steps instead of the O(N ) queries
ment [43]. As the number of bits in the secret string increases, required classically where N is the number of items in the
the probability of correctly guessing the string becomes less database. Grover’s algorithm is a fundamental demonstration
dependent on the type of disorder and more reliant on the cen- of quantum computing’s potential for solving certain search
ter and spread of the disorder [44]. The classical algorithm, and optimization problems faster than classical computers. It
on the other hand, becomes inefficient for long strings, even was proposed by Lov Grover in 1996 [50]. Since classical al-
in a noiseless scenario [45]. Overall, the Bernstein-Vazirani gorithms for NP-complete problems (nondeterministic poly-
algorithm outperforms classical algorithms in most cases. nomial time problems) require exponentially many steps, and
VOLUME 4, 2023 11

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4
 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2024.3362955

Shafique et al.: Quantum Computing: Circuits, Algorithms, and Applications

(a) Single-Qubit NOT gate on one qubit example

(b) Single-Qubit Hadamard gate on one qubit example

(c) Multiple-Qubit SWAP gate on two qubits example

(d) Multiple-Qubit Controlled-X gate on two qubits example

(e) Lower order gate matrices applied on higher order qubit states

(f) Two Qubits entanglement example

12 VOLUME 4, 2023

FIGURE 7: Basic circuits of quantum systems.

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4
 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2024.3362955

Shafique et al.: Quantum Computing: Circuits, Algorithms, and Applications

Grover’s algorithm provides at most a quadratic speedup; this [65]. These novel quantum algorithms utilize parameterized
suggests that Grover’s algorithm by itself will not provide quantum circuits, often referred to as Parameterized Quan-
polynomial-time solutions for NP-complete problems [51]. tum Circuits (PQCs) or Quantum Neural Networks (QNNs)
[66] in deep learning applications. Similar to classical deep
VI. DESIGNING HYBRID CLASSICAL-QUANTUM learning, QNN parameters are optimized with respect to a
MACHINE LEARNING (ML) MODELS cost function through black-box optimization heuristics [67]
Machine learning (ML) involves creating models or algo- or gradient-based methods [68]. This optimization aims to
rithms that are capable of learning patterns within datasets. facilitate the learning of data representations.
ML includes regression, classification, segmentation, object
detection, etc. After learning from a dataset, these models be- C. HYBRID MACHINE LEARNING
come capable of predicting outcomes for unseen data. Orig- Quantum processors are still fairly small and noisy, therefore
inally, analytical approaches were used in Machine learn- to improve machine learning performance effectively, NISQ
ing [52]; however, recently, heuristic-based methods have processors will need to work with classical co-processors in
become more dominant due to the abundance of data and hybrid mode. Scalability, training time, and model accuracy
computational resources [53]. These methods have achieved are the key factors in the field of deep learning [69] and these
substantial success in the field of deep learning [54], [55]. are the limitations of classical deep learning methods. These
Deep learning methods develop data representation in the issues can also be addressed with the combination of classical
form of high-dimensional vectors using parameterized layers. and quantum processors. The abstract model of the classical-
Optimization techniques are used to fine-tune these parame- quantum deep neural network is shown in Figure 8.
ters and fit the deep learning model.
Concurrently, academia and industry have witnessed a
growing interest in quantum computing [34]. Quantum com-
puting utilizes quantum mechanics principles such as super-
position, entanglement, and quantum parallelism to improve
the performance of classical machine learning algorithms.
Quantum computers exhibit these novel behaviors that are of-
ten challenging to simulate using classical computers. These
novel behaviors of quantum computers make them capable
of solving certain problems more efficiently as compared to FIGURE 8: Abstract structure for hybrid classical-quantum
classical systems. Quantum computers have the potential to model. ψ represents quantum model (QNN) parameters
accelerate tasks such as optimization [56] and cryptanalysis
[57]. Liu et al. have proposed a hybrid quantum-classical con-
volutional neural network (QCCNN), similar to convolu-
A. FIRST GENERATION OF QUANTUM MACHINE tional neural networks (CNNs) in classical deep learning but
LEARNING adapted to quantum computing to improve the process of fea-
Quantum computing is capable of enhancing the machine ture mapping [70]. Another study [71] has proposed a hybrid
learning design process. A pivotal development leading to the quantum-classical graph convolutional network (QGCNN)
utilization of quantum computers in machine learning is their for learning high-energy physics data. The proposed frame-
ability to speed up linear algebra operations exponentially work has shown an advantage over classical multilayer per-
as state space grows. These QML algorithms, based on ceptron and convolutional neural networks in the aspect of
quantum-accelerated linear algebra techniques, form the first the number of parameters with comparable accuracy. This
generation of QML algorithms, addressing the set of super- paper Liang et al. [72] have proposed the idea of a hybrid
vised and unsupervised learning tasks [11], [58]–[60]. These quantum–classical neural network with deep residual learn-
algorithms offer solutions that are faster compared to their ing (Res-HQCNN). They have designed a residual block
classical counterparts for certain problems. For example, the structure with a quantum neural network and the correspond-
quantum technique for solving systems of linear equations ing training algorithm.
[61] is utilized for classification problems such as perceptron
or linear regression training. VII. QUANTUM SIMULATORS
Quantum simulators are used to test and debug quantum cir-
B. SECOND GENERATION OF QUANTUM MACHINE cuits with realistic noise models. Following are five notable
LEARNING quantum simulators:
With the advent of NISQ processors, the second genera-
tion of QML has emerged [62]–[64]. Differing from the A. QISKIT AER
first generation, this new trend in QML relies on heuristic Qiskit Aer [73] is a high-performance quantum computing
methods due to the increased computational capabilities of simulator provided by IBM. It provides interfaces to run
quantum hardware particularly in the field of deep learning quantum circuits with various noise models. It also includes
VOLUME 4, 2023 13

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4
 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2024.3362955

Shafique et al.: Quantum Computing: Circuits, Algorithms, and Applications

state vector and density matrix simulators for realistic simu- a temperature of 10-20mK is required using a state-of-the-
lations. It can also utilize graphical processing units (GPUs) art dilution refrigerator, the heart of quantum computing
to improve simulation performance. infrastructure.
Qubits are created with superconducting materials. At ex-
B. QUTIP (QUANTUM TOOLBOX IN PYTHON) tremely low (near absolute zero) temperatures, the supercon-
QuTiP [74] is an open-source quantum computing framework ducting circuits act as superconducting material as they carry
for Python. It is designed for simulating the dynamics of open current at low electrical resistivity. A Josephson junction is
quantum systems. This framework depends on the Scipy, an assembly of two weakly coupled superconductors with
Cython, and Numpy packages. In addition, Matplotlib is used a thin layer of insulator. At superconducting temperature,
for graphical output. QuTiP is user-friendly and free of any electric current flows through the Josephson junction [79]
licensing fees, therefore, it is considered suitable for learning with no applied voltage. To generate a qubit, a linear ca-
quantum computing in the classroom. pacitor made of a superconducting material and an insulator
is tied together to superconducting wires with Josephson
C. CIRQ junction in the conducting loop to form an artificial atom
which holds the qubit. At near absolute zero temperature,
Cirq [75] is an open-source interface designed for program-
the qubit has no thermal energy in the surroundings, so it
ming quantum computers. It is a Python software library
stays in a stable state. The qubits are then entangled using a
developed for creating and simulating quantum circuits on
microwave pulse with a range depending upon the strength
quantum processors. Cirq also includes a simulator to run
of the magnetic field of the qubit or its resonance frequency.
quantum algorithms for testing and debugging.
The quantum computer setup consists of logical and physical
qubits. Physical qubits are the actual quantum bits imple-
D. PROJECTQ
mented in quantum hardware. They are the primary building
ProjectQ [76] is an open-source programming interface for blocks of quantum computers, typically found in quantum
quantum computers. It supports compilation framework for systems like superconducting circuits, photons, or trapped
various types of quantum hardware such as IBM Quantum ions. Physical qubits are susceptible to errors and noise due
Experience chip, Azure Quantum, AWS Braket, AQT de- to the external environment. Several physical qubits or the
vices, or ion-trap quantum (IonQ) devices. ProjectQ also actual quantum hardware are grouped together to form a
includes a simulator for testing and debugging quantum single logical qubit. Logical qubits are a higher-level concept
algorithms. that represents quantum information that is protected against
noise and errors. The logical qubits are used for error and
E. PYQUIL noise detection.
pyQuil [77] is a Python library for writing quantum programs The superconducting quantum computing cooling setup
using Quil, the quantum instruction language developed at mainly consists of the following components and is shown
Rigetti Computing, Inc. It comes with a quantum virtual in Figure 9,
machine (QVM) for simulating and testing quantum circuits. • Dilution refrigerator (Cryostat with pulse tube cooler)
• Gas handling system with gas/liquid mixture pumps,
VIII. QUANTUM COMPUTING DEPLOYMENT valves and controls
REQUIREMENTS • Liquid Nitrogen Dewars
Quantum computing promises to outperform classical com- 3 4
• He and He storage tanks
puting by solving certain problems. Though still in the de- • Cryo compressor
velopment phase, there are many approaches to quantum • Cryostat control panel
computing namely superconducting, photonic, trapped ion,
neutral atoms, and quantum dots. Superconducting quantum
computing is the widely used type of quantum supercomput-
ing. Due to highly customized end use, and the limitations of
the resources such as Quantum hardware, quantum experts
believe that the design, delivery, and deployment of this
type of computing can be challenging. The topology that is
the most studied and vetted of the superconducting circuit
approaches is the nearest-neighbor cavity coupled transmon
qubits [78].
Quantum computing circuits operate at superconducting FIGURE 9: Quantum computing cooling infrastructure lay-
temperatures. The qubits are sensitive to environmental in- out.
terference such as electromagnetic and thermal fluctuations;
hence the system needs to be maintained at near absolute zero The quantum qubits need to be stored at near absolute
temperature. Depending on the type of quantum computing, zero temperatures for which a mixture of 3 He and 4 He along
14 VOLUME 4, 2023

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4
 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2024.3362955

Shafique et al.: Quantum Computing: Circuits, Algorithms, and Applications

with nitrogen is used in the dilution refrigerator [80]. Liquid tasks exponentially faster than classical computing. Quantum
nitrogen is used for 80K cooling and the He mixture is to computing can be used to enhance algorithms and discover
achieve 2K cooling. The liquid nitrogen is also used to cool structures and patterns efficiently. In this section, we will dive
the 3 He return from the system and to remove any air impu- into the applications of quantum computing in more detail.
rities. Due to its sensitive nature, the presence of any AC and Quantum computing applications are summarized in Table 2.
DC magnetic fields may be checked in the room housing the
cryostat prior to its installation. Helium and its isotope used A. CRYPTOGRAPHY
in the cryogenic cooling process may potentially leak and
Cryptography involves the study of secure communication,
pose issues with the depletion of oxygen in space. A helium
and quantum computing has the potential to undermine nu-
leak detector with oxygen deficiency hazard alarm sensors
merous encryption techniques employed in classical cryp-
is recommended to be installed in quantum computing lab
tography, including RSA and elliptic curve cryptography.
space.
However, quantum computing can also be used to create new
The gas handling system (GHS) is an essential element of a
cryptographic protocols that are resistant to quantum attacks.
quantum cooling system. It comprises a cabinet that contains
For example, it can be used for secure key distribution over
the 3 He and 4 He mixture reservoir, pumps for circulation,
quantum channels, which have advantages over classical
valves, chilled water, liquid nitrogen connection, and controls
channels in terms of detecting eavesdropping [81]. Addition-
(mostly pneumatic). The operation of GHS is considerably
ally, Quantum cryptography schemes can protect wireless
noisy, so it is recommended to place it in separate rooms
sensor networks from attacks by hackers [92]. Quantum com-
alongside the cryostat compressors. Also, vibration isolators
puting has emerged as a compelling complement to classical
are recommended under the Dilution refrigeration assembly.
technologies for applications in security and communications
The ambient space of the quantum setup as shown in Figure 9
[93]. In a nutshell, quantum computing offers promising so-
may be maintained at 68-70F dry bulb and 30-50% relative
lutions for cryptanalysis as well as for enhancing the security
humidity levels.
and efficiency of cryptographic systems.
Unlike conventional high-performance computing (HPC)
using graphics processing units (GPUs) and central process-
ing units (CPUs) with thermal design power (TDP) of more B. OPTIMIZATION
than 1000W, a superconducting quantum system relies on Optimization is the process of finding the best solution to a
low power consumption. For operational reliability of the given problem from a set of solutions. Quantum computers
quantum computer, clean power from a UPS source is gen- can efficiently solve certain optimization problems that are
erally recommended. The nominal voltage ranges are 120V- complicated tasks for classical computers. Quantum algo-
480V. Depending on the size of the qubit system, the Cryo rithms such as Grover search, quantum phase estimation,
compressor may need a dedicated chiller to provide lower- quantum annealing, quantum approximate optimization al-
temperature cooling water generally at a higher delta > 30F gorithm, and variational quantum eigensolver can be used
across the compressor. to solve optimization problems [94]. Additionally, quan-
The requirements for connectivity are generally scalable tum computing can be applied in flight path optimization
and flexible as the need and user community grows over time. within the aerospace engineering domain, where quantum
A microwave amplifier, signal synthesizers, a multi-channel computing can tackle computational challenges and improve
arbitrary pulse sequencer, 1 gigabyte of DDR3 SDRAM, and performance over classical algorithms [95].
a gigabit Ethernet interface for high-speed data upload are
generally required for a qubit setup.
C. CHEMISTRY
There are many moving components in the cryogenics
of the Quantum Infrastructure. Most of the components Quantum computing has numerous applications in pharma-
do not require frequent maintenance, however, proactive ceuticals and drug discovery. It can be used for generative
preventative maintenance can ensure the longevity of the chemistry and quantum chemistry simulations [96]. By using
entire quantum computing system. The parts are manufac- quantum computing, drug companies can potentially save
tured by multiple vendors; therefore assembly, integration, time and money by accelerating the drug discovery process,
and commissioning of the entire system can be challenging. leading to a more efficient and productive pharmaceutical
The quantum computing operators need to learn and obtain industry [97]. Quantum computing can also reduce costs and
hands-on experience to ensure the optimum functioning of time in drug development by decreasing the number of neces-
the setup. sary biochemical experiments [98]. Quantum computing has
been used in protein structure prediction, molecular docking,
IX. APPLICATIONS OF QUANTUM COMPUTING and quantum simulation [99]. Although current quantum
Quantum computing has the potential to transform many devices are still susceptible to external noise and error, but
areas including astrophysics, aerospace, pharmaceuticals, hybrid quantum-classical techniques are well suited for drug
drug discovery, artificial intelligence, cybersecurity, and se- discovery and development.
cure communication. It has the potential to perform certain
VOLUME 4, 2023 15

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4
 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2024.3362955

Shafique et al.: Quantum Computing: Circuits, Algorithms, and Applications

TABLE 2: Summary of quantum applications with algorithm/model used.

Task Algorithm/ Model Study Used for
Secure Key Distribu- Quantum channels Rana et al. [81] Secure communication and detection of eavesdropping.
tion
Quantum Variational Classifier Adhikary et al. [82] Fisher’s Iris, Sonar, and Wisconsin’s Breast Cancer (WBC)
Classification datasets.
Quantum inspired Nearest Sergioli et al. [83] Idiopathic pulmonary fibrosis (IPF) dataset.
Mean Classifier
Quantum Neural Network Xiong et al. [84] Image classification using MNIST digits and CIFAR-10
datasets.
Quantum Neural Network Ngo et al. [85] Li-ion battery degradation dataset.
Regression
Continuous Variable - Kanimozhi et al. [86] Surface plasmon resonance sensor dataset.
Quantum Neural Network
(CV-QNN)
Reinforcement Hybrid-Quantum Proximal Rainjonneau et al. [87] Satellite path planning and task decision using multiple satel-
Learning Policy Optimization lites dataset.
Hybrid Quantum Deep Q Net- Bar et al. [88] Synthetic robot navigation dataset.
work
Dimension Reduction Quantum PCA Dri et al. [89] Synthetic financial interest rate dataset.
Image Generation Quantum–Classical GAN Tsang et al. [90] High-resolution image generation using MNIST digits and
(QGANs) Fashion datasets.
Contrastive Learning Quantum SimCLR Model Jaderberg et al. [91] Image processing using CIFAR-10 dataset.

D. MACHINE LEARNING X. CONCLUSIONS AND FUTURE OF QUANTUM

Quantum computing has shown potential to improve machine COMPUTING
learning model accuracy. Quantum machine learning algo- This paper has offered a thorough exploration of the field
rithms such as quantum neural networks, have been applied of quantum computing, making its complexities more un-
in the context of image classification for iris, sonar, breast derstandable. We have covered the fundamental concepts of
cancer, idiopathic pulmonary fibros, CIFAR-10 and MNIST quantum mechanics and the evolving quantum computing
datasets [82]–[84]. These quantum models have shown ben- landscape with practical applications. This paper has dis-
efits over classical models, including reduced training time cussed noisy intermediate-scale quantum (NISQ) along with
and improved model accuracy. While there is still a need for quantum computing fundamentals, such as qubits’ notations,
further development of quantum hardware, quantum machine quantum superposition, quantum entanglement, quantum in-
learning holds great potential for improving the efficiency of terference, and quantum noise. We have elaborated building
classical machine learning methods. blocks of quantum systems, such as quantum gates, quantum
measurements, and basic quantum circuits. We have pro-
E. ENERGY vided an overview of various quantum algorithms, such as
Quantum computing has various applications in the field the Deutsch-Jozsa algorithm, Bernstein-Vazirani algorithm,
of energy. One area of application is in power and en- Shor’s algorithm, and Grover’s algorithm. We have discussed
ergy systems, where quantum computing algorithms like the design of hybrid classical-quantum machine learning
Grover’s algorithm can be applied to solve problems more algorithms. We have further discussed the deployment re-
efficiently than traditional algorithms. Another application quirements of superconducting quantum computing, the most
area is battery profiling, where quantum computing can be prevalent form of quantum supercomputing.
used to model the degradation profile of the battery [85]. Quantum computing has the ability to solve problems
Additionally, quantum computing can contribute to simulat- that are beyond the capabilities of any classical computer,
ing linear and nonlinear dynamics in fusion energy science a capability referred to as quantum supremacy. Achieving
applications, which can help in understanding wave-particle true quantum supremacy would be a major milestone in the
interactions and nonlinear plasma dynamics [100]. field of quantum computing. However, with such powerful
technology, it also raises issues such as privacy and security.
F. FINANCE Researchers will need to consider these issues as quantum
Quantum computing has numerous applications in finance. computing technology grows.
It can be used for financial interest rate modeling [89]. It In conclusion, we recognize the remarkable progress in
has also been applied to develop financial models, such as quantum computing and its potential to address complicated
churn prediction and credit risk assessment, where it has problems with efficiency more than classical computers.
demonstrated better performance compared to traditional This tutorial is intended to be a valuable resource for those
methods [101]. Quantum computing offers significant ben- beginning their journey into quantum computing and for
efits in terms of computational speed and accuracy, making it researchers to stay informed about this exciting field. The
a valuable tool in the finance field [102]. future of quantum computing is promising, and we can
16 VOLUME 4, 2023

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4
 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2024.3362955

Shafique et al.: Quantum Computing: Circuits, Algorithms, and Applications

expect to see significant breakthroughs in scientific research, [21] T. Van der Sar, Z. Wang, M. Blok, H. Bernien, T. Taminiau, D. Toyli,
engineering, and technology in the coming years. D. Lidar, D. Awschalom, R. Hanson, and V. Dobrovitski, “Decoherence-
protected quantum gates for a hybrid solid-state spin register,” Nature,
vol. 484, no. 7392, pp. 82–86, 2012.
ACKNOWLEDGMENTS [22] P. W. Shor, “Scheme for reducing decoherence in quantum computer
This study acknowledges the University of Engineering and memory,” Physical review A, vol. 52, no. 4, p. R2493, 1995.
[23] P. Shor, “Proceedings of 37th conference on foundations of computer
Technology (UET), Lahore for the support and affiliation science, burlington, 1996,” 1996.
with the author Muhammad Ali Shafique. [24] E. Knill, R. Laflamme, and W. H. Zurek, “Resilient quantum computa-
tion,” Science, vol. 279, no. 5349, pp. 342–345, 1998.
[25] J. Preskill, “Reliable quantum computers,” Proceedings of the Royal
REFERENCES Society of London. Series A: Mathematical, Physical and Engineering
[1] R. Barends, J. Kelly, A. Megrant, A. Veitia, D. Sank, E. Jeffrey, T. C. Sciences, vol. 454, no. 1969, pp. 385–410, 1998.
White, J. Mutus, A. G. Fowler, B. Campbell et al., “Superconducting [26] F. Gaitan, Quantum error correction and fault tolerant quantum
quantum circuits at the surface code threshold for fault tolerance,” Nature, computing. CRC Press, 2008.
vol. 508, no. 7497, pp. 500–503, 2014. [27] D. A. Lidar and T. A. Brun, Quantum error correction. Cambridge
[2] J. Biamonte, P. Wittek, N. Pancotti, P. Rebentrost, N. Wiebe, and university press, 2013.
S. Lloyd, “Quantum machine learning,” Nature, vol. 549, no. 7671, pp. [28] G. G. La Guardia, “Quantum error correction,” Quantum science and
195–202, 2017. technology. Cham: Springer, 2020.
[3] S. Debnath, N. M. Linke, C. Figgatt, K. A. Landsman, K. Wright, and [29] A. G. Fowler, M. Mariantoni, J. M. Martinis, and A. N. Cleland, “Surface
C. Monroe, “Demonstration of a small programmable quantum computer codes: Towards practical large-scale quantum computation,” Physical
with atomic qubits,” Nature, vol. 536, no. 7614, pp. 63–66, 2016. Review A, vol. 86, no. 3, p. 032324, 2012.
[4] N. Ofek, A. Petrenko, R. Heeres, P. Reinhold, Z. Leghtas, B. Vlastakis, [30] F. Arute, K. Arya, R. Babbush, D. Bacon, J. C. Bardin, R. Barends,
Y. Liu, L. Frunzio, S. Girvin, L. Jiang et al., “Extending the lifetime of R. Biswas, S. Boixo, F. G. Brandao, D. A. Buell et al., “Quantum
a quantum bit with error correction in superconducting circuits,” Nature, supremacy using a programmable superconducting processor,” Nature,
vol. 536, no. 7617, pp. 441–445, 2016. vol. 574, no. 7779, pp. 505–510, 2019.
[5] P. W. Shor, “Polynomial-time algorithms for prime factorization and [31] H.-S. Zhong, H. Wang, Y.-H. Deng, M.-C. Chen, L.-C. Peng, Y.-H. Luo,
discrete logarithms on a quantum computer,” SIAM review, vol. 41, no. 2, J. Qin, D. Wu, X. Ding, Y. Hu et al., “Quantum computational advantage
pp. 303–332, 1999. using photons,” Science, vol. 370, no. 6523, pp. 1460–1463, 2020.
[6] L. G. Sandor Imre, Advanced quantum communications: an engineering [32] J. Chow, O. Dial, and J. Gambetta, “Ibm quantum breaks the 100-qubit
approach. Wiley-IEEE Press, 2013. processor barrier,” IBM Research Blog, vol. 2, 2021.
[7] R. L. Rivest, A. Shamir, and L. Adleman, “A method for obtaining digital [33] L. S. Madsen, F. Laudenbach, M. F. Askarani, F. Rortais, T. Vincent,
signatures and public-key cryptosystems,” Communications of the ACM, J. F. Bulmer, F. M. Miatto, L. Neuhaus, L. G. Helt, M. J. Collins
vol. 21, no. 2, pp. 120–126, 1978. et al., “Quantum computational advantage with a programmable photonic
[8] P. Shor, “Polynomial-time algorithms for prime factorization and discrete processor,” Nature, vol. 606, no. 7912, pp. 75–81, 2022.
logarithms on a quantum computer. foundations of computer science.” [34] J. Preskill, “Quantum computing in the nisq era and beyond,” Quantum,
Conference Publications, 1997. vol. 2, p. 79, 2018.
[9] J. Proos and C. Zalka, “Shor’s discrete logarithm quantum algorithm for [35] B. Schumacher, “Quantum coding,” Phys. Rev. A, vol. 51, pp. 2738–
elliptic curves,” 2004. 2747, Apr 1995. [Online]. Available: https://link.aps.org/doi/10.1103/
[10] T. Albash and D. A. Lidar, “Adiabatic quantum computation,” Reviews PhysRevA.51.2738
of Modern Physics, vol. 90, no. 1, p. 015002, 2018. [36] P. A. M. Dirac, “A new notation for quantum mechanics,” Mathematical
[11] S. Lloyd, M. Mohseni, and P. Rebentrost, “Quantum algorithms Proceedings of the Cambridge Philosophical Society, vol. 35, no. 3, p.
for supervised and unsupervised machine learning,” arXiv preprint 416–418, 1939.
arXiv:1307.0411, 2013. [37] J. Solem and L. Biedenharn, “Understanding geometrical phases in
[12] T. F. Rønnow, Z. Wang, J. Job, S. Boixo, S. V. Isakov, D. Wecker, J. M. quantum mechanics: An elementary example,” Foundations of physics,
Martinis, D. A. Lidar, and M. Troyer, “Defining and detecting quantum vol. 23, no. 2, pp. 185–195, 1993.
speedup,” science, vol. 345, no. 6195, pp. 420–424, 2014. [38] V. Vedral and M. B. Plenio, “Basics of quantum computation,” Progress
[13] I. L. Chuang, R. Laflamme, P. W. Shor, and W. H. Zurek, “Quantum in Quantum Electronics, vol. 22, no. 1, pp. 1–39, jan 1998. [Online].
computers, factoring, and decoherence,” Science, vol. 270, no. 5242, pp. Available: https://doi.org/10.1016%2Fs0079-6727%2898%2900004-4
1633–1635, 1995. [39] A. Steane, “Quantum computing,” Reports on Progress in Physics,
[14] W. G. Unruh, “Maintaining coherence in quantum computers,” Physical vol. 61, no. 2, pp. 117–173, feb 1998. [Online]. Available: https:
Review A, vol. 51, no. 2, p. 992, 1995. //doi.org/10.1088%2F0034-4885%2F61%2F2%2F002
[15] B. Georgeot and D. L. Shepelyansky, “Quantum chaos border for quan- [40] P. Aradyamath, N. Naghabhushana, and R. Ujjinimatad, “Quantum
tum computing,” Physical Review E, vol. 62, no. 3, p. 3504, 2000. computing concepts with deutsch jozsa algorithm,” JOIV: International
[16] G. Kalai, “The argument against quantum computers,” Quantum, Journal on Informatics Visualization, vol. 3, no. 1, pp. 59–68, 2019.
probability, logic: The work and influence of Itamar Pitowsky, pp. 399– [41] R. De, R. Moberly, C. Beery, J. Juybari, and K. Sundqvist, “Multi-qubit
422, 2020. size-hopping deutsch-jozsa algorithm with qubit reordering for secure
[17] R. Babbush, J. R. McClean, M. Newman, C. Gidney, S. Boixo, and quantum key distribution,” in 2021 IEEE International Conference on
H. Neven, “Focus beyond quadratic speedups for error-corrected quan- Quantum Computing and Engineering (QCE). IEEE, 2021, pp. 473–
tum advantage,” PRX Quantum, vol. 2, no. 1, p. 010103, 2021. 474.
[18] L. M. Vandersypen, M. Steffen, G. Breyta, C. S. Yannoni, M. H. Sher- [42] A. N. Oliveira, E. V. de Oliveira, A. C. Santos, and C. J. Villas-Bôas,
wood, and I. L. Chuang, “Experimental realization of shor’s quantum “Quantum algorithms in ibmq experience: Deutsch-jozsa algorithm,”
factoring algorithm using nuclear magnetic resonance,” Nature, vol. 414, arXiv preprint arXiv:2109.07910, 2021.
no. 6866, pp. 883–887, 2001. [43] A. Gupta, P. Ghosh, K. Sen, and U. Sen, “Effects of noise on performance
[19] S. Gulde, M. Riebe, G. P. Lancaster, C. Becher, J. Eschner, H. Häffner, of bernstein-vazirani algorithm,” arXiv preprint arXiv:2305.19745, 2023.
F. Schmidt-Kaler, I. L. Chuang, and R. Blatt, “Implementation of the [44] P. Fernández and M. A. Martin-Delgado, “Homomorphic encription of
deutsch–jozsa algorithm on an ion-trap quantum computer,” Nature, vol. the k= 2 bernstein-vazirani algorithm,” arXiv preprint arXiv:2303.17426,
421, no. 6918, pp. 48–50, 2003. 2023.
[20] L. DiCarlo, J. M. Chow, J. M. Gambetta, L. S. Bishop, B. R. Johnson, [45] X. Zhou, D. Qiu, and L. Lou, “Distributed exact quantum algo-
D. Schuster, J. Majer, A. Blais, L. Frunzio, S. Girvin et al., “Demonstra- rithms for bernstein-vazirani and search problems,” arXiv preprint
tion of two-qubit algorithms with a superconducting quantum processor,” arXiv:2303.10670, 2023.
Nature, vol. 460, no. 7252, pp. 240–244, 2009. [46] X. Cheng, Y. Xu, K. Wang, Y. Zhang, B. Li, and Z. Zhang, “Lightweight
and flexible hardware implementation of authenticated encryption al-
gorithm simon-galois/counter mode,” International Journal of Circuit
Theory and Applications.
VOLUME 4, 2023 17

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4
 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2024.3362955

Shafique et al.: Quantum Computing: Circuits, Algorithms, and Applications

[47] X. Bonnetain, “Tight bounds for simon’s algorithm,” in Progress [73] Qiskit contributors, “Qiskit: An open-source framework for quantum
in Cryptology–LATINCRYPT 2021: 7th International Conference on computing,” 2023.
Cryptology and Information Security in Latin America, Bogotá, [74] J. J. P.D. Nation. Qutip. [Online]. Available: https://qutip.org/
Colombia, October 6–8, 2021, Proceedings 7. Springer, 2021, pp. 3– [75] C. Developers, “Cirq,” Zenodo, 2023.
23. [76] Projectq - an open source software framework for quantum computing.
[48] P. W. Shor, “Algorithms for quantum computation: discrete logarithms [Online]. Available: https://github.com/ProjectQ-Framework/ProjectQ
and factoring,” in Proceedings 35th annual symposium on foundations of [77] R. S. Smith, M. J. Curtis, and W. J. Zeng, “A practical quantum instruc-
computer science. Ieee, 1994, pp. 124–134. tion set architecture,” 2016.
[49] J. P. Buhler, H. W. Lenstra, and C. Pomerance, “Factoring integers with [78] X. Jin, K. Cicak, Z. Parrott, S. Kotler, F. Lecocq, J. Teufel, J. Aumentado,
the number field sieve,” in The development of the number field sieve. E. Kapit, and R. Simmonds, “Versatile parametric coupling between two
Springer, 1993, pp. 50–94. statically decoupled transmon qubits,” arXiv preprint arXiv:2305.02907,
[50] L. K. Grover, “A fast quantum mechanical algorithm for database search,” 2023.
in Proceedings of the twenty-eighth annual ACM symposium on Theory [79] R. Newrock, C. Lobb, U. Geigenmüller, and M. Octavio, “The two-
of computing, 1996, pp. 212–219. dimensional physics of josephson-junction arrays,” SOLID STATE
[51] M. A. Nielsen and I. L. Chuang, Quantum computation and quantum PHYSICS-NEW YORK-ACADEMIC PRESS-, vol. 54, pp. 266–512,
information. Cambridge university press, 2010. 2000.
[52] K. P. Murphy, Machine learning: a probabilistic perspective. MIT press, [80] H. Dang, R. Zha, J. Tan, T. Zhang, J. Li, N. Li, B. Zhao, Y. Zhao, H. Tan,
2012. and R. Xue, “Investigations on a 3.3 k four-stage stirling-type pulse tube
[53] Y. LeCun, Y. Bengio, and G. Hinton, “Deep learning,” nature, vol. 521, cryocooler. part b: Experimental verifications,” Cryogenics, vol. 105, p.
no. 7553, pp. 436–444, 2015. 103015, 2020.
[54] A. Krizhevsky, “Advances in neural information processing systems,” [81] H. Rana and N. Verma, “Enhanced quantum key distribution us-
(No Title), p. 1097, 2012. ing hybrid channels and natural random numbers,” arXiv preprint
[55] I. Goodfellow, Y. Bengio, and A. Courville, Deep learning. MIT press, arXiv:2007.14298, 2020.
2016. [82] S. Adhikary, S. Dangwal, and D. Bhowmik, “Supervised learning with
[56] E. Farhi, J. Goldstone, and S. Gutmann, “A quantum approximate opti- a quantum classifier using multi-level systems,” Quantum Information
mization algorithm,” arXiv preprint arXiv:1411.4028, 2014. Processing, vol. 19, pp. 1–12, 2020.
[57] P. Shor, “35th annual symposium on foundations of computer science, [83] G. Sergioli, G. Russo, E. Santucci, A. Stefano, S. E. Torrisi, S. Palmucci,
1994 proceedings,” 1994. C. Vancheri, and R. Giuntini, “Quantum-inspired minimum distance
[58] S. Lloyd, M. Mohseni, and P. Rebentrost, “Quantum principal component classification in a biomedical context,” International Journal of Quantum
analysis,” Nature Physics, vol. 10, no. 9, pp. 631–633, 2014. Information, vol. 16, no. 08, p. 1840011, 2018.
[59] P. Rebentrost, M. Mohseni, and S. Lloyd, “Quantum support vector [84] H. Xiong, X. Duan, Y. Yu, J. Zhang, and H. Yin, “Image classifica-
machine for big data classification,” Physical review letters, vol. 113, tion based on quantum machine learning,” in 2023 5th International
no. 13, p. 130503, 2014. Conference on Intelligent Control, Measurement and Signal Processing
[60] I. Kerenidis and A. Prakash, “Quantum recommendation systems,” arXiv (ICMSP), 2023, pp. 891–895.
preprint arXiv:1603.08675, 2016. [85] A. P. Ngo, N. Le, H. T. Nguyen, A. Eroglu, and D. T. Nguyen, “A
[61] B. Duan, J. Yuan, C.-H. Yu, J. Huang, and C.-Y. Hsieh, “A survey on quantum neural network regression for modeling lithium-ion battery
hhl algorithm: From theory to application in quantum machine learning,” capacity degradation,” in 2023 IEEE Green Technologies Conference
Physics Letters A, vol. 384, no. 24, p. 126595, 2020. (GreenTech), 2023, pp. 164–168.
[62] L. Zhou, S.-T. Wang, S. Choi, H. Pichler, and M. D. Lukin, “Quantum [86] T. Kanimozhi, S. Sridevi, M. Valliammai, J. Mohanraj, and V. Kumar,
approximate optimization algorithm: Performance, mechanism, and im- “Quantum regression model for the prediction of surface plasmon res-
plementation on near-term devices,” Physical Review X, vol. 10, no. 2, p. onance sensor behaviour,” in 2022 Workshop on Recent Advances in
021067, 2020. Photonics (WRAP). IEEE, 2022, pp. 1–2.
[63] S. McArdle, T. Jones, S. Endo, Y. Li, S. C. Benjamin, and X. Yuan, “Vari- [87] S. Rainjonneau, I. Tokarev, S. Iudin, S. Rayaprolu, K. Pinto, D. Lemtiuzh-
ational ansatz-based quantum simulation of imaginary time evolution,” nikova, M. Koblan, E. Barashov, M. Kordzanganeh, M. Pflitsch, and
npj Quantum Information, vol. 5, no. 1, p. 75, 2019. A. Melnikov, “Quantum algorithms applied to satellite mission planning
[64] Z. Jiang, J. McClean, R. Babbush, and H. Neven, “Majorana loop for earth observation,” IEEE Journal of Selected Topics in Applied Earth
stabilizer codes for error mitigation in fermionic quantum simulations,” Observations and Remote Sensing, vol. 16, pp. 7062–7075, 2023.
Physical Review Applied, vol. 12, no. 6, p. 064041, 2019. [88] N. F. Bar, H. Yetis, and M. Karakose, “An approach based on quantum
[65] M. Mohseni, P. Read, H. Neven, S. Boixo, V. Denchev, R. Babbush, reinforcement learning for navigation problems,” in 2022 International
A. Fowler, V. Smelyanskiy, and J. Martinis, “Commercialize quantum Conference on Data Analytics for Business and Industry (ICDABI),
technologies in five years,” Nature, vol. 543, no. 7644, pp. 171–174, 2022, pp. 593–597.
2017. [89] E. Dri, A. Aita, T. Fioravanti, G. Franco, E. Giusto, G. Ranieri, D. Corbel-
[66] H. Chen, L. Wossnig, S. Severini, H. Neven, and M. Mohseni, “Uni- letto, and B. Montrucchio, “Towards an end-to-end approach for quantum
versal discriminative quantum neural networks,” Quantum Machine principal component analysis,” in 2023 IEEE International Conference on
Intelligence, vol. 3, pp. 1–11, 2021. Quantum Computing and Engineering (QCE), vol. 02, 2023, pp. 1–6.
[67] G. Verdon, M. Broughton, J. R. McClean, K. J. Sung, R. Babbush, [90] S. L. Tsang, M. T. West, S. M. Erfani, and M. Usman, “Hybrid quan-
Z. Jiang, H. Neven, and M. Mohseni, “Learning to learn with quan- tum–classical generative adversarial network for high-resolution image
tum neural networks via classical neural networks,” arXiv preprint generation,” IEEE Transactions on Quantum Engineering, vol. 4, pp. 1–
arXiv:1907.05415, 2019. 19, 2023.
[68] R. Sweke, F. Wilde, J. Meyer, M. Schuld, P. K. Fährmann, B. Meynard- [91] B. Jaderberg, L. W. Anderson, W. Xie, S. Albanie, M. Kiffner, and
Piganeau, and J. Eisert, “Stochastic gradient descent for hybrid quantum- D. Jaksch, “Quantum self-supervised learning,” Quantum Science and
classical optimization,” Quantum, vol. 4, p. 314, 2020. Technology, vol. 7, no. 3, p. 035005, 2022.
[69] M. Broughton, G. Verdon, T. McCourt, A. J. Martinez, J. H. Yoo, S. V. [92] A. Montanaro, “Quantum algorithms: an overview,” npj Quantum
Isakov, P. Massey, R. Halavati, M. Y. Niu, A. Zlokapa et al., “Tensorflow Information, vol. 2, no. 1, pp. 1–8, 2016.
quantum: A software framework for quantum machine learning,” arXiv [93] V. Hassija, V. Chamola, A. Goyal, S. S. Kanhere, and N. Guizani, “Forth-
preprint arXiv:2003.02989, 2020. coming applications of quantum computing: peeking into the future,” IET
[70] J. Liu, K. H. Lim, K. L. Wood, W. Huang, C. Guo, and H.-L. Huang, “Hy- Quantum Communication, vol. 1, no. 2, pp. 35–41, 2020.
brid quantum-classical convolutional neural networks,” Science China [94] Y. Wang, J. E. Kim, and K. Suresh, “Opportunities and challenges of
Physics, Mechanics & Astronomy, vol. 64, no. 9, p. 290311, 2021. quantum computing for engineering optimization,” Journal of Computing
[71] S. Y.-C. Chen, T.-C. Wei, C. Zhang, H. Yu, and S. Yoo, “Hy- and Information Science in Engineering, vol. 23, no. 6, 2023.
brid quantum-classical graph convolutional network,” arXiv preprint [95] H. Makhanov, K. Setia, J. Liu, V. Gomez-Gonzalez, and G. Jenaro-
arXiv:2101.06189, 2021. Rabadan, “Quantum computing applications for flight trajectory opti-
[72] Y. Liang, W. Peng, Z.-J. Zheng, O. Silvén, and G. Zhao, “A hybrid mization,” arXiv preprint arXiv:2304.14445, 2023.
quantum–classical neural network with deep residual learning,” Neural
Networks, vol. 143, pp. 133–147, 2021.
18 VOLUME 4, 2023

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4
 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2024.3362955

Shafique et al.: Quantum Computing: Circuits, Algorithms, and Applications

[96] A. Pyrkov, A. Aliper, D. Bezrukov, Y.-C. Lin, D. Polykovskiy, P. Kamya, [99] M. Avramouli, I. K. Savvas, A. Vasilaki, and G. Garani, “Unlocking
F. Ren, and A. Zhavoronkov, “Quantum computing for near-term applica- the potential of quantum machine learning to advance drug discovery,”
tions in generative chemistry and drug discovery,” Drug Discovery Today, Electronics, vol. 12, no. 11, p. 2402, 2023.
p. 103675, 2023. [100] H. M. Gray and K. Terashi, “Quantum computing applications in future
[97] V. Mahesh and S. Shijo, “Accelerating drug discovery with quantum colliders,” Frontiers in Physics, vol. 10, p. 864823, 2022.
computing,” Evolution and Applications of Quantum Computing, pp. [101] K. A. Tychola, T. Kalampokas, and G. A. Papakostas, “Quantum machine
175–181, 2023. learning—an overview,” Electronics, vol. 12, no. 11, p. 2379, 2023.
[98] P.-H. Wang, J.-H. Chen, Y.-Y. Yang, C. Lee, and Y. J. Tseng, “Recent [102] Y.-J. Chang, M.-F. Sie, S.-W. Liao, and C.-R. Chang, “The prospects
advances in quantum computing for drug discovery and development,” of quantum computing for quantitative finance and beyond,” IEEE
IEEE Nanotechnology Magazine, 2023. Nanotechnology Magazine, 2023.

VOLUME 4, 2023 19

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4
 This article has been accepted for publication in IEEE Access. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/ACCESS.2024.3362955

Shafique et al.: Quantum Computing: Circuits, Algorithms, and Applications

ALI SHAFIQUE is a Ph.D. candidate in the De- IMRAN LATIF is currently serving as the Chief
partment of Computer Science at Kansas State Operations Officer at the U.S. Department of En-
University. He also holds the position of research ergy, Office of Science, at Brookhaven National
assistant in the ISCAAS laboratory. Ali Shafique Laboratory where he leads the infrastructure op-
earned his Bachelor of Science degree from the erations of high-performance computing centers
University of Engineering and Technology, La- supporting global cutting-edge research collabora-
hore, where he completed his studies between tions in physics, life sciences, quantum and com-
2011 and 2015. He obtained his Master’s degree putational sciences, and artificial intelligence. Mr.
from the same University from 2016 to 2017. Imran has over 20 years of leadership experience
Shafique has developed a keen interest in the field delivering large-scale engineering, construction,
of efficient Machine Learning in resource-constrained systems, He is com- and sustainability projects with great success for top companies including
mitted to conducting research in this area with the aim of advancing the IBM, Microsoft, Google, Wall Street District, and many others. Mr. Imran
current knowledge and understanding of these fields. is also a keynote speaker and has delivered expert talks at various tech
and entrepreneur conferences worldwide, including DCD, BICSI, Super
Computing, HEPIX, and CHEP. He completed his masters in engineering
from The City University of New York. He holds professional engineering
licenses in the states of New York and Texas along with several professional
ARSLAN MUNIR (M’09, SM’17) is currently an certifications including CEM, PMP, DCEP, CDCDP and LEED. He is also
Associate Professor in the Department of Com- a mentor for the U.S. Federal Government-funded internship programs
puter Science at Kansas State University. He was where he works with students from Ivy league universities on leading-
a postdoctoral research associate in the Electrical edge scientific research projects. Mr. Imran is presently spearheading global
and Computer Engineering (ECE) department at research initiatives with leading academia and big tech companies on data
Rice University, Houston, Texas, USA from May center energy conservation. He is the founder and CEO of New York-based
2012 to June 2014. He received his M.A.Sc. in engineering design and consulting firm providing professional services to
ECE from the University of British Columbia Fortune 500 clients within retail, healthcare, telecom, energy, and industrial
(UBC), Vancouver, Canada, in 2007 and his Ph.D. sectors.
in ECE from the University of Florida (UF),
Gainesville, Florida, USA, in 2012. From 2007 to 2008, he worked as
a software development engineer at Mentor Graphics Corporation in the
Embedded Systems Division. Munir’s current research interests include em-
bedded and cyber-physical systems, secure and trustworthy systems, parallel
computing, artificial intelligence, and computer vision. Munir received many
academic awards including the doctoral fellowship from Natural Sciences
and Engineering Research Council (NSERC) of Canada. He earned gold
medals for best performance in electrical engineering, gold medals and
academic roll of honor for securing rank one in pre-engineering provincial
examinations (out of approximately 300,000 candidates). He is a Senior
Member of IEEE.

20 VOLUME 4, 2023

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4