Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

Implementation of quantum gate operations using neural networks

Shahri A. Zabihi

Published May 1, 2017• 0 citations
OpenAlex

Abstract

All quantum circuits are designed using different quantum gates, which can be decomposed into elementary gates. The number of elementary gates in a quantum circuit is called the gate count. Typically, there is a direct correspondence between the gate count and the complexity of a quantum circuit. Quantum systems are generally very fragile, and a quantum bit (qubit) can lose its super-position state very easily. This process is called decoherence. As such, when implementing a quantum operation, it becomes necessary to minimize the gate count as much as possible. This thesis focuses on two important applications where reducing the gate count is very significant. The first is quantum error correction (QEC). In a quantum error correction code (QECC), one information qubit is encoded with two or more auxiliary qubits to form a logical/encoded qubit. Oftentimes, when performing gate operations on logical qubits, decoding is required, which opens the system to decoherence. The aim here is to design encoded quantum gates in such a way that the decoding process is no longer needed for implementing gate operations on QEC circuits. The second focus is a controlled-NOT (CNOT) gate operation between uncoupled (remote) qubits. In order to implement a gate operation in a linear nearest neighbor (LNN) architecture, the qubits that are not neighbors need to be brought adjacent to each other before a gate operation can be performed between them. The LNN architecture is significant because most physical implementations of a practical quantum computer use this layout. Here, the aim is to implement a CNOT gate between remote qubits in an LNN architecture without bringing the qubits adjacent to each other, thereby tremendously reducing the gate count. This research employed a neural network approach based on gradient descent technique to reduce the gate count.

Related papers

Reducing CNOT count in quantum Fourier transform for the linear nearest-neighbor architecture

Byeongyong Park, Doyeol Ahn

Physical limitations of quantum hardware often necessitate nearest-neighbor (NN) architecture. When synthesizing quantum circuits using the basic gate library, which consists of CNOT and single-qubit gates, CNOT gates are required to convert a quantum circuit into one suitable for an NN architecture. In the basic gate library, CNOT gates are considered the primary cost of quantum circuits due to their higher error rates and longer execution times compared to single-qubit gates. In this paper, we propose a new linear NN (LNN) circuit design for quantum Fourier transform (QFT), one of the most versatile subroutines in quantum algorithms. Our LNN QFT circuit has only about 40% of the number of CNOT gates compared to previously known LNN QFT circuits. Subsequently, we input both our QFT circuits and conventional QFT circuits into the Qiskit transpiler to construct QFTs on IBM quantum computers, which necessitate NN architectures. Consequently, our QFT circuits demonstrate a substantial advantage over conventional QFT circuits in terms of the number of CNOT gates. This outcome implies that the proposed LNN QFT circuit design could serve as a novel foundation for developing QFT circuits implemented in quantum hardware that demands NN architecture.

arXiv:2208.14249•Scientific Reports•May 27, 2023•34 citations

An efficient conversion of quantum circuits to a linear nearest neighbor architecture

Yuichi Hirata, Masaki Nakanishi, Shigeru Yamashita +1 more

Several promising implementations of quantum computation rely on a Linear Nearest Neighbor (LNN) architecture, which arranges quantum bits on a line, and allows neighbor interactions only. Therefore, several specific circuits have been designed on an LNN architecture. However, a general and efficient conversion method for an arbitrary circuit has not been established. Therefore, this paper gives an efficient conversion technique to convert quantum circuits to an LNN architecture. When a quantum circuit is converted to an LNN architecture, the objective is to reduce the size of the additional circuit added by the conversion and the time complexity of the conversion. The proposed method requires less additional circuitry and time complexity compared with naive techniques. To develop the method, we introduce two key theorems that may be interesting on their own. In addition, the proposed method also achieves less overhead than some known circuits designed from scratch on an LNN architecture.

Quantum Information and Computation•Jan 1, 2011•79 citations

An Efficient Method to Convert Arbitrary Quantum Circuits to Ones on a Linear Nearest Neighbor Architecture

Yuichi Hirata, Masaki Nakanishi, Shigeru Yamashita +1 more

A variety of quantum circuits have been designed.Most of them assume that arbitrary pairs of qubits can interact.However, several promising implementations of quantum computation rely on a Linear Nearest Neighbor (LNN) architecture, which arranges quantum bits on a line, and allows neighbor interactions only.Therefore, several specific circuits have been designed on an LNN architecture.However, a general and efficient conversion technique for an arbitrary circuit has not been established.Therefore, this paper gives an efficient method that converts an arbitrary quantum circuit to one on an LNN architecture.Our method achieves small overhead and time complexity compared with naive techniques.To develop the method, we introduce two key theorems that may be interesting on their own.In addition, our method also achieves smaller overhead for some known circuits designed on an LNN architecture.

International Conference on Quantum, Nano and Micro Technologies (ICQNM)•Feb 1, 2009•73 citations

Nearest-Neighbor and Fault-Tolerant Quantum Circuit Implementation

Laxmidhar Biswal, Chandan Bandyopadhyay, Anupam Chattopadhyay +3 more

The quest of achieving higher computing performance is driving the research on quantum computing, which is reporting new milestones almost on a daily basis. For practical quantum circuit design, fault tolerance is an essential condition. This is achieved by mapping the target functions into the Clifford+T group of elementary quantum gates. Furthermore, the application of error-correcting codes in quantum circuits requires the quantum gates to be formed between adjacent Qubits. In this work, we improve the state-of-the-art quantum circuit design by addressing both of the above challenges. First, we propose a novel mapping of Multiple-Control Toffoli (MCT) gates to Clifford+T group gates, which achieves lower gate count compared to earlier work. Secondly, we show a generic way to convert any Clifford+T circuit into a nearest neighbor one. We validate the efficacy of our approach with detailed experimental studies.

IEEE International Symposium on Multiple-Valued Logic (ISMVL)•May 1, 2016•18 citations

Quantum-error correction on linear-nearest-neighbor qubit arrays

Austin G. Fowler, Charles D. Hill, Lloyd C. L. Hollenberg

A quantum circuit implementing 5-qubit quantum-error correction on a linear-nearest-neighbor architecture is described. The canonical decomposition is used to construct fast and simple gates that incorporate the necessary swap operations allowing the circuit to achieve the same depth as the current least depth circuit. Simulations of the circuit's performance when subjected to discrete and continuous errors are presented. The relationship between the error rate of a physical qubit and that of a logical qubit is investigated with emphasis on determining the concatenated error correction threshold.

arXiv:quant-ph/0311116•Physical Review A•Apr 20, 2004•61 citations

Multi-Controlled Quantum Gates in Linear Nearest Neighbor

Ben Zindorf, Sougato Bose

Multi-controlled single-target (MC) gates are some of the most crucial building blocks for varied quantum algorithms. How to implement them optimally is thus a pivotal question. To answer this question in an architecture-independent manner, and to get a worst-case estimate, we should look at a linear nearest-neighbor (LNN) architecture, as this can be embedded in almost any qubit connectivity. Motivated by the above, here we describe a method which implements MC gates using no more than $\sim 4k+8n$ CNOT gates -- up-to $60\%$ reduction over state-of-the-art -- while allowing for complete flexibility to choose the locations of $n$ controls, the target, and a dirty ancilla out of $k$ qubits. More strikingly, in case $k \approx n$, our upper bound is $\sim 12n$ -- the best known for unrestricted connectivity -- and if $n = 1$, our upper bound is $\sim 4k$ -- the best known for a single long-range CNOT gate over $k$ qubits -- therefore, if our upper bound can be reduced, then the cost of one or both of these simpler versions of MC gates will be immediately reduced accordingly. In practice, our method provides circuits that tend to require fewer CNOT gates than our upper bound for almost any given instance of MC gates.

arXiv:2506.00695•May 31, 2025•7 citations
All quantum circuits are designed using different quantum gates, which can be
decomposed into elementary gates. The number of elementary gates in a quantum circuit is
called the gate count. Typically, there is a direct correspondence between the gate count and the
complexity of a quantum circuit. Quantum systems are generally very fragile, and a quantum bit
(qubit) can lose its super-position state very easily. This process is called decoherence. As such,
when implementing a quantum operation, it becomes necessary to minimize the gate count as
much as possible.
This thesis focuses on two important applications where reducing the gate count is very
significant. The first is quantum error correction (QEC). In a quantum error correction code
(QECC), one information qubit is encoded with two or more auxiliary qubits to form a
logical/encoded qubit. Oftentimes, when performing gate operations on logical qubits, decoding
is required, which opens the system to decoherence. The aim here is to design encoded quantum
gates in such a way that the decoding process is no longer needed for implementing gate
operations on QEC circuits. The second focus is a controlled-NOT (CNOT) gate operation
between uncoupled (remote) qubits. In order to implement a gate operation in a linear nearest
neighbor (LNN) architecture, the qubits that are not neighbors need to be brought adjacent to
each other before a gate operation can be performed between them. The LNN architecture is
significant because most physical implementations of a practical quantum computer use this
layout. Here, the aim is to implement a CNOT gate between remote qubits in an LNN
architecture without bringing the qubits adjacent to each other, thereby tremendously reducing
the gate count. This research employed a neural network approach based on gradient descent
technique to reduce the gate count.
Physical limitations of quantum hardware often necessitate nearest-neighbor (NN) architecture. When synthesizing quantum circuits using the basic gate library, which consists of CNOT and single-qubit gates, CNOT gates are required to convert a quantum circuit into one suitable for an NN architecture. In the basic gate library, CNOT gates are considered the primary cost of quantum circuits due to their higher error rates and longer execution times compared to single-qubit gates. In this paper, we propose a new linear NN (LNN) circuit design for quantum Fourier transform (QFT), one of the most versatile subroutines in quantum algorithms. Our LNN QFT circuit has only about 40% of the number of CNOT gates compared to previously known LNN QFT circuits. Subsequently, we input both our QFT circuits and conventional QFT circuits into the Qiskit transpiler to construct QFTs on IBM quantum computers, which necessitate NN architectures. Consequently, our QFT circuits demonstrate a substantial advantage over conventional QFT circuits in terms of the number of CNOT gates. This outcome implies that the proposed LNN QFT circuit design could serve as a novel foundation for developing QFT circuits implemented in quantum hardware that demands NN architecture.
Several promising implementations of quantum computation rely on a Linear Nearest Neighbor (LNN) architecture, which arranges quantum bits on a line, and allows neighbor interactions only. Therefore, several specific circuits have been designed on an LNN architecture. However, a general and efficient conversion method for an arbitrary circuit has not been established. Therefore, this paper gives an efficient conversion technique to convert quantum circuits to an LNN architecture. When a quantum circuit is converted to an LNN architecture, the objective is to reduce the size of the additional circuit added by the conversion and the time complexity of the conversion. The proposed method requires less additional circuitry and time complexity compared with naive techniques. To develop the method, we introduce two key theorems that may be interesting on their own. In addition, the proposed method also achieves less overhead than some known circuits designed from scratch on an LNN architecture.
A variety of quantum circuits have been designed.Most of them assume that arbitrary pairs of qubits can interact.However, several promising implementations of quantum computation rely on a Linear Nearest Neighbor (LNN) architecture, which arranges quantum bits on a line, and allows neighbor interactions only.Therefore, several specific circuits have been designed on an LNN architecture.However, a general and efficient conversion technique for an arbitrary circuit has not been established.Therefore, this paper gives an efficient method that converts an arbitrary quantum circuit to one on an LNN architecture.Our method achieves small overhead and time complexity compared with naive techniques.To develop the method, we introduce two key theorems that may be interesting on their own.In addition, our method also achieves smaller overhead for some known circuits designed on an LNN architecture.
The quest of achieving higher computing performance is driving the research on quantum computing, which is reporting new milestones almost on a daily basis. For practical quantum circuit design, fault tolerance is an essential condition. This is achieved by mapping the target functions into the Clifford+T group of elementary quantum gates. Furthermore, the application of error-correcting codes in quantum circuits requires the quantum gates to be formed between adjacent Qubits. In this work, we improve the state-of-the-art quantum circuit design by addressing both of the above challenges. First, we propose a novel mapping of Multiple-Control Toffoli (MCT) gates to Clifford+T group gates, which achieves lower gate count compared to earlier work. Secondly, we show a generic way to convert any Clifford+T circuit into a nearest neighbor one. We validate the efficacy of our approach with detailed experimental studies.
A quantum circuit implementing 5-qubit quantum-error correction on a linear-nearest-neighbor architecture is described. The canonical decomposition is used to construct fast and simple gates that incorporate the necessary swap operations allowing the circuit to achieve the same depth as the current least depth circuit. Simulations of the circuit's performance when subjected to discrete and continuous errors are presented. The relationship between the error rate of a physical qubit and that of a logical qubit is investigated with emphasis on determining the concatenated error correction threshold.
Multi-controlled single-target (MC) gates are some of the most crucial building blocks for varied quantum algorithms. How to implement them optimally is thus a pivotal question. To answer this question in an architecture-independent manner, and to get a worst-case estimate, we should look at a linear nearest-neighbor (LNN) architecture, as this can be embedded in almost any qubit connectivity. Motivated by the above, here we describe a method which implements MC gates using no more than ∼4k+8n\sim 4k+8n∼4k+8n CNOT gates -- up-to 60%60\%60% reduction over state-of-the-art -- while allowing for complete flexibility to choose the locations of nnn controls, the target, and a dirty ancilla out of kkk qubits. More strikingly, in case k≈nk \approx nk≈n, our upper bound is ∼12n\sim 12n∼12n -- the best known for unrestricted connectivity -- and if n=1n = 1n=1, our upper bound is ∼4k\sim 4k∼4k -- the best known for a single long-range CNOT gate over kkk qubits -- therefore, if our upper bound can be reduced, then the cost of one or both of these simpler versions of MC gates will be immediately reduced accordingly. In practice, our method provides circuits that tend to require fewer CNOT gates than our upper bound for almost any given instance of MC gates.