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

IEEE International Conference on Consumer Electronics - Taiwan (ICCE-Taiwan)

A Quantum State Preparation Method Based on Gram-Schmidt Decomposition

Su‐Ling Lee, Chien‐Cheng Tseng

Published Jul 16, 2025• 0 citations
Publisher

Abstract

Quantum state preparation (QSP) refers to the process of initializing a quantum circuit into a specific quantum state, which serves as the starting point for several quantum algorithms. In this article, a Gram-Schmidt decomposition (GSD) method is proposed to address the QSP problem. The method involves the following five steps: First, the first column vector of the identity matrix is replaced with the unit-norm data vector to construct an initial matrix. Second, the initial matrix is decomposed into the product of a unitary matrix and an upper triangular matrix using the GSD method. Third, the transpile tool from IBM Qiskit is employed to design a quantum circuit, ensuring that the input and output state vectors correspond to the relationship defined by the obtained unitary matrix. Fourth, when the input state vector of the quantum circuit is set as the zero vector, the output state vector corresponds to the desired data vector to be prepared. Finally, a numerical example using the Iris flower dataset is provided to demonstrate the effectiveness of the proposed method.

DOI
10.1109/icce-taiwan66881.2025.11208090

Related papers

A General Quantum Circuit Design Method for Preparation of W State

Su‐Ling Lee, Chien‐Cheng Tseng

State preparation is a fundamental task in quantum computation and quantum communication, as it initializes a quantum circuit into a desired quantum state. Notable examples of quantum states include the W state, Greenberger–Horne–Zeilinger (GHZ) state, and Dicke state. In this article, we propose a general method for designing quantum circuits to generate the W state. The approach begins by replacing the second column of the identity matrix with the target W state vector, forming an initial matrix. This matrix is then factorized using QR decomposition into a unitary matrix Q and an upper triangular matrix R. A quantum circuit that realizes the unitary matrix Q is subsequently obtained using the IBM Qiskit transpiler. The transformation performed by the circuit corresponds to matrix-vector multiplication with Q, and the desired W state emerges as the output when the input is the all-zero quantum state. Experimental results for W state generation with three qubits are presented to demonstrate the effectiveness of the proposed method.

IEEE Global Conference on Consumer Electronics (GCCE)•Sep 23, 2025•0 citations

An Implementation Method of Digital FIR Filter Using Quantum State Preparation Circuit

Chien‐Cheng Tseng, Su‐Ling Lee

Digital finite-impulse-response (FIR) filters are a cornerstone of digital signal processing (DSP) and modern communication systems, making their implementation via quantum computation a compelling area of study. Although recent methods have used orthogonal triangular decomposition (OTD) to realize FIR filters through quantum elementary gates, these approaches rely heavily on matrix QR decomposition and IBM Qiskit's transpilation tools, resulting in significant computational overhead. To address these inefficiencies, this paper proposes a new implementation using quantum state preparation (QSP) circuits. By employing only rotation y gate and multi-controlled rotation y gates, the proposed method determines circuit parameters through closed-form solutions derived directly from the input signal and filter coefficient vectors. Experimental results for a digital mean filter demonstrate the effectiveness of this QSP-based approach, showing the reduced implementation complexity compared to the conventional OTD method.

IEEE VTS Asia Pacific Wireless Communications Symposium (APWCS)•Aug 1, 2026•0 citations

Quantum algorithm for vector set orthogonal normalization and matrix QR decomposition with polynomial speedup

Zi-Ming Li, Yu-xi Liu

Vector set orthogonal normalization and matrix QR decomposition are fundamental problems in matrix analysis with important applications in many fields. We know that the Gram-Schmidt process is a widely used method to solve these two problems. However, existing methods, including the Gram-Schmidt process, have problems of high complexity, scaling $O({N}^{3})$ in the system dimension $N$, which leads to difficulties when calculating large-scale or ill-conditioned problems. With the development of quantum information processing, a series of quantum algorithms have been proposed, providing advantages and speedups over classical algorithms in many fields. In this paper we propose quantum algorithms to solve these two problems based on the idea of the Gram-Schmidt process and quantum phase estimation. The complexity of proposed quantum algorithms is theoretically and numerically analyzed. We find that our algorithms provide polynomial acceleration over the best-known quantum algorithms and a potential polynomial acceleration over the best-known classical algorithms on these two problems, scaling $O({N}^{2}\mathrm{poly}(logN))$ in the dimension $N$ of the system when ignoring the complexity of state readout or $O({N}^{2}\mathrm{poly}(logN))$ when considering the complexity of state readout.

arXiv:2412.19090•Physical Review A•Aug 11, 2025•3 citations

A Closed-Form Parameter Determination Method for Quantum State Preparation Circuit

Su‐Ling Lee, Chien‐Cheng Tseng

Designing a quantum circuit that prepares userspecific states is a fundamental task in quantum computation. This involves loading user-provided data into the initial superposition state of a quantum register, enabling further processing by quantum algorithms. To address this state preparation problem, a quantum circuit utilizing rotation Y gates and controlled rotation Y gates has been previously proposed. However, the method for determining the rotation angles of these Y gates has not been detailed. This paper introduces a closed-form method for computing these angle parameters based on user-provided data. Numerical examples demonstrate the ease of use of the proposed method, which is implemented using IBM’s quantum information software kit (Qiskit). Additionally, real-world feature data from iris flowers is used to show that the feature data can be accurately loaded into the quantum register, facilitating tasks such as machine learning, dimensionality reduction, or pattern recognition through quantum computation.

International Symposium on Intelligent Signal Processing and Communication Systems (ISPACS)•Dec 10, 2024•5 citations

Quantum Gram-Schmidt processes and their application to efficient state readout for quantum algorithms

Kaining Zhang, Min-Hsiu Hsieh, Liu Liu +1 more

Many quantum algorithms that claim speedup over their classical counterparts only generates quantum states as solutions instead of their final classical description. The additional step to decode quantum states into classical vectors normally will destroy the quantum advantage in most scenarios because all existing tomographic methods require runtime that is polynomial with respect to the state dimension. In this work, we present an efficient readout protocol that yields the classical vector form of the generated state, so it will achieve the end-to-end advantage for those quantum algorithms. Our protocol suits the case in which the output state lies in the row space of the input matrix, of rank $r$, that is stored in the quantum random access memory. The quantum resources for decoding the state in ${\ensuremath{\ell}}^{2}$ norm with $\ensuremath{\epsilon}$ error require $\mathrm{poly}(r,1/\ensuremath{\epsilon})$ copies of the output state and $\mathrm{poly}(r,{\ensuremath{\kappa}}^{r},1/\ensuremath{\epsilon})$ queries to the input oracles, where $\ensuremath{\kappa}$ is the condition number of the input matrix. With our readout protocol, we completely characterize the end-to-end resources for quantum linear equation solvers and quantum singular value decomposition. One of our technical tools is an efficient quantum algorithm for performing the Gram-Schmidt orthonormal procedure, which we believe will be of independent interest.

arXiv:2004.06421•Physical Review Research•Nov 1, 2021•15 citations

Dimensionality Reduction Using Quantum Computation and Data Transformation

Su‐Ling Lee, Chien‐Cheng Tseng

In this article, we present a dimensionality reduction method based on data transformation and quantum computation. The proposed approach is implemented using a cascade quantum circuit comprising three key stages: quantum state preparation (QSP), quantum data transformation (QDT), and quantum measurement. The main steps of the method are as follows: First, the QSP circuit is designed using elementary quantum gates to encode the given data into a quantum register, enabling efficient data transformation. Next, the QDT circuit processes the encoded data, compressing it into a more compact representation with fewer significant features. This study explores two types of data transformations: dataindependent transforms (DIT), commonly used in digital signal compression, such as the discrete cosine transform (DCT), discrete Walsh transform (DWT), and discrete Haar transform (DHT); and data-dependent transforms (DDT), which leverage statistical properties of the data, such as the Karhunen-Loeve transform (KLT) which is also known as principal component analysis (PCA). Finally, quantum measurement is performed to extract the transformed data, where small-valued coefficients are discarded to achieve dimensionality reduction. To validate the effectiveness of the proposed method, experiments are conducted using real-world Fisher's iris flower data set.

International Symposium on Computer, Consumer and Control•Jun 27, 2025•0 citations
Quantum state preparation (QSP) refers to the process of initializing a quantum circuit into a specific quantum state, which serves as the starting point for several quantum algorithms. In this article, a Gram-Schmidt decomposition (GSD) method is proposed to address the QSP problem. The method involves the following five steps: First, the first column vector of the identity matrix is replaced with the unit-norm data vector to construct an initial matrix. Second, the initial matrix is decomposed into the product of a unitary matrix and an upper triangular matrix using the GSD method. Third, the transpile tool from IBM Qiskit is employed to design a quantum circuit, ensuring that the input and output state vectors correspond to the relationship defined by the obtained unitary matrix. Fourth, when the input state vector of the quantum circuit is set as the zero vector, the output state vector corresponds to the desired data vector to be prepared. Finally, a numerical example using the Iris flower dataset is provided to demonstrate the effectiveness of the proposed method.
State preparation is a fundamental task in quantum computation and quantum communication, as it initializes a quantum circuit into a desired quantum state. Notable examples of quantum states include the W state, Greenberger–Horne–Zeilinger (GHZ) state, and Dicke state. In this article, we propose a general method for designing quantum circuits to generate the W state. The approach begins by replacing the second column of the identity matrix with the target W state vector, forming an initial matrix. This matrix is then factorized using QR decomposition into a unitary matrix Q and an upper triangular matrix R. A quantum circuit that realizes the unitary matrix Q is subsequently obtained using the IBM Qiskit transpiler. The transformation performed by the circuit corresponds to matrix-vector multiplication with Q, and the desired W state emerges as the output when the input is the all-zero quantum state. Experimental results for W state generation with three qubits are presented to demonstrate the effectiveness of the proposed method.
Digital finite-impulse-response (FIR) filters are a cornerstone of digital signal processing (DSP) and modern communication systems, making their implementation via quantum computation a compelling area of study. Although recent methods have used orthogonal triangular decomposition (OTD) to realize FIR filters through quantum elementary gates, these approaches rely heavily on matrix QR decomposition and IBM Qiskit's transpilation tools, resulting in significant computational overhead. To address these inefficiencies, this paper proposes a new implementation using quantum state preparation (QSP) circuits. By employing only rotation y gate and multi-controlled rotation y gates, the proposed method determines circuit parameters through closed-form solutions derived directly from the input signal and filter coefficient vectors. Experimental results for a digital mean filter demonstrate the effectiveness of this QSP-based approach, showing the reduced implementation complexity compared to the conventional OTD method.
Vector set orthogonal normalization and matrix QR decomposition are fundamental problems in matrix analysis with important applications in many fields. We know that the Gram-Schmidt process is a widely used method to solve these two problems. However, existing methods, including the Gram-Schmidt process, have problems of high complexity, scaling O(N3)O({N}^{3})O(N3) in the system dimension NNN, which leads to difficulties when calculating large-scale or ill-conditioned problems. With the development of quantum information processing, a series of quantum algorithms have been proposed, providing advantages and speedups over classical algorithms in many fields. In this paper we propose quantum algorithms to solve these two problems based on the idea of the Gram-Schmidt process and quantum phase estimation. The complexity of proposed quantum algorithms is theoretically and numerically analyzed. We find that our algorithms provide polynomial acceleration over the best-known quantum algorithms and a potential polynomial acceleration over the best-known classical algorithms on these two problems, scaling O(N2poly(logN))O({N}^{2}\mathrm{poly}(logN))O(N2poly(logN)) in the dimension NNN of the system when ignoring the complexity of state readout or O(N2poly(logN))O({N}^{2}\mathrm{poly}(logN))O(N2poly(logN)) when considering the complexity of state readout.
Designing a quantum circuit that prepares userspecific states is a fundamental task in quantum computation. This involves loading user-provided data into the initial superposition state of a quantum register, enabling further processing by quantum algorithms. To address this state preparation problem, a quantum circuit utilizing rotation Y gates and controlled rotation Y gates has been previously proposed. However, the method for determining the rotation angles of these Y gates has not been detailed. This paper introduces a closed-form method for computing these angle parameters based on user-provided data. Numerical examples demonstrate the ease of use of the proposed method, which is implemented using IBM’s quantum information software kit (Qiskit). Additionally, real-world feature data from iris flowers is used to show that the feature data can be accurately loaded into the quantum register, facilitating tasks such as machine learning, dimensionality reduction, or pattern recognition through quantum computation.
Many quantum algorithms that claim speedup over their classical counterparts only generates quantum states as solutions instead of their final classical description. The additional step to decode quantum states into classical vectors normally will destroy the quantum advantage in most scenarios because all existing tomographic methods require runtime that is polynomial with respect to the state dimension. In this work, we present an efficient readout protocol that yields the classical vector form of the generated state, so it will achieve the end-to-end advantage for those quantum algorithms. Our protocol suits the case in which the output state lies in the row space of the input matrix, of rank rrr, that is stored in the quantum random access memory. The quantum resources for decoding the state in ${\ensuremath{\ell}}^{2}$ norm with $\ensuremath{\epsilon}$ error require $\mathrm{poly}(r,1/\ensuremath{\epsilon})$ copies of the output state and $\mathrm{poly}(r,{\ensuremath{\kappa}}^{r},1/\ensuremath{\epsilon})$ queries to the input oracles, where $\ensuremath{\kappa}$ is the condition number of the input matrix. With our readout protocol, we completely characterize the end-to-end resources for quantum linear equation solvers and quantum singular value decomposition. One of our technical tools is an efficient quantum algorithm for performing the Gram-Schmidt orthonormal procedure, which we believe will be of independent interest.
In this article, we present a dimensionality reduction method based on data transformation and quantum computation. The proposed approach is implemented using a cascade quantum circuit comprising three key stages: quantum state preparation (QSP), quantum data transformation (QDT), and quantum measurement. The main steps of the method are as follows: First, the QSP circuit is designed using elementary quantum gates to encode the given data into a quantum register, enabling efficient data transformation. Next, the QDT circuit processes the encoded data, compressing it into a more compact representation with fewer significant features. This study explores two types of data transformations: dataindependent transforms (DIT), commonly used in digital signal compression, such as the discrete cosine transform (DCT), discrete Walsh transform (DWT), and discrete Haar transform (DHT); and data-dependent transforms (DDT), which leverage statistical properties of the data, such as the Karhunen-Loeve transform (KLT) which is also known as principal component analysis (PCA). Finally, quantum measurement is performed to extract the transformed data, where small-valued coefficients are discarded to achieve dimensionality reduction. To validate the effectiveness of the proposed method, experiments are conducted using real-world Fisher's iris flower data set.