Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1808

Integrating Quantum Algorithms with

Gravitational-Wave Metrology for
Enhanced Signal Detection
Vaidik A Sharma
Birla Institute of Technology and Science Pilani, India

Abstract:- This study explores the integration of sensors. The integration of quantum technologies aims to
quantum algorithms, specifically Grover's algorithm, enhance sensitivity and create scalable solutions for
with quantum metrology to enhance the efficiency and understanding mysterious phenomena in the universe.
sensitivity of gravitational-wave detection. By combining
quantum matched filtering with precise parameter II. BACKGROUND
estimation techniques, the research aims to optimize
sensor networks for the identification of gravitational The combination of Quantum algorithm for
waves. This integrated approach leverages the strengths gravitationalwave matched filtering and quantum metrology
of quantum superposition and entanglement to improve presents a comprehensive solution for selecting the most
signal detection, reduce noise, and strategically place effective quantum sensor network for detecting gravitational
sensors. The findings demonstrate significant waves. By integrating these two techniques, we can improve
improvements in the sensitivity and accuracy of the network's performance in terms of efficiency and
gravitational wave measurements, highlighting the sensitivity, ultimately advancing our understanding of
potential of quantum technologies to revolutionize elusive cosmic phenomena.
observational astronomy and enhance our
understanding of the universe.  Swift Signal Detection: Utilizing Grover's algorithm, the
Quantum Algorithm for Gravitational-Wave Matched
Keywords:- Quantum Algorithms, Gravitational-Wave Filtering accelerates signal detection in noisy data,
Detection, LIGO/Virgo Data Analysis, Matched Filtering, significantly reducing identification time.
Quantum Metrology, Parameter Estimation.  Precise Parameter Estimation: Quantum metrology
complements matched filtering by providing accurate
I. INTRODUCTION measurements of crucial parameters like frequency,
amplitude, and phase associated with gravitational wave
Exploring the transformative potential of quantum signals. This precision enhances the overall accuracy of
technologies has opened promising avenues in quantum the detection process.
sensing and communication. Quantum computers, although  Heightened Sensitivity: The synergy between quantum
still developing, offer solutions to complex problems metrology and the quantum algorithm amplifies sensor
beyond classical computers' reach. Alongside quantum sensitivity, enabling the detection of even faint
computing, Noisy Intermediate-Scale Quantum (NISQ) gravitational wave signals that may elude classical
devices have emerged as significant frontiers. Despite sensors.
inherent errors, NISQ computers exhibit unique capabilities  Noise Reduction: Quantum metrology contributes to
in various applications like optimization and cryptography. noise reduction within the network, ensuring high-
In quantum sensing, they show promise in enhancing quality data collection necessary for accurate template
detection precision, including potential roles in gravitational matching in the filtering process.
wave detection. Gravitational waves, originating from  Strategic Sensor Placement: Quantum metrology assists
massive cosmic events, present challenges for study. While in determining optimal sensor placement by estimating
traditional detectors like LIGO and Virgo are effective, parameters related to the gravitational wave source, such
they're costly and have limitations, prompting exploration of as its location and propagation direction, thus optimizing
more efficient methods, like integrating quantum sensors signal detection efficiency.
into networks. Research investigates integrating quantum  Scalability: This integrated approach facilitates the
sensing and communication in NISQ sensor networks for efficient scalability of quantum sensor networks to
gravitational wave detection. Advanced quantum algorithms accommodate diverse research scenarios and evolving
such as Quantum Algorithm for Gravitational-Wave scientific requirements.
Matched Filtering aim to leverage quantum properties for
improved detection. Quantum metrology parameter
estimation holds promise for navigation, timekeeping, and
environmental monitoring. The study addresses practical
implications and challenges of deploying quantum-enhanced

IJISRT24MAY1808 www.ijisrt.com 2189

 Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1808

The integration of advanced quantum techniques holds  Variational Parameter Estimation and Quantum
immense potential in revolutionizing gravitational wave Metrology
detection. By combining innovative quantum algorithms and Variational parameter estimation is a powerful
precise parameter estimation, we can tailor sensor networks technique in quantum computation,[2] particularly in the
for improved performance, accelerating signal detection, context of metrology. Quantum metrology leverages the
refining measurements, and minimizing noise. Ultimately, principles of quantum superposition and entanglement to
this advancement brings us closer to unlocking the mysteries achieve measurements with higher precision than classical
of the universe with greater accuracy and efficiency. methods allow. Variational circuits, parameterized by angles
𝜃, can be used to prepare quantum states and perform
III. QUANTUM SIMULATION AND METROLOGY measurements. The goal is to find the optimal parameters
TECHNIQUE that minimize a loss function 𝐿(𝜃) and produce the desired
quantum state.[5],[6]
 Hamiltonian Simulation and Time Evolution
At the heart of quantum simulation lies the Mathematically, the variational parameter estimation
Hamiltonian operator 𝐻, which encapsulates the total energy process can be formulated as an optimization problem:
of a quantum system. The time evolution of a quantum state
|𝜓(𝑡)⟩ governed by the Hamiltonian 𝐻 is described by the 𝜃 ∗ = argmin𝐿(𝜃).
Schrödinger equation: 𝜃

𝑑 Quantum metrology techniques use the optimized

𝑖ℏ |𝜓(𝑡)⟩ = 𝐻|𝜓(𝑡)⟩ parameters to enhance the accuracy of parameter
𝑑𝑡 estimation.[2] The Fisher information (𝐹) quantifies the
This equation highlights the dynamics of quantum sensitivity of the quantum state to variations in the
states as they evolve over time under the influence of 𝐻. parameter 𝜃. In bra-ket notation, it is given by:
Quantum computers excel at simulating this time evolution, 2
enabling the study of quantum systems that are challenging 𝑑
|⟨𝜓𝑘 | 𝜓(𝜃)⟩|
or impossible to simulate using classical methods [1]. The 𝑑𝜃
𝐹(𝜃) = ∑
ability to simulate quantum systems holds promise for tasks 𝑝𝑘
𝑘
such as understanding chemical reactions, optimizing
materials, simulating condensed matter systems and 𝑑
Where 𝑝𝑘 is the probability of outcome 𝑘 and 𝑑𝜃 𝜓(𝜃)
gravitational wave astronomy [11].
is the derivative of the quantum state with respect to 𝜃. The
 The Time Block Method: Mitigating Simulation Errors Fisher information sets a fundamental limit on how
In quantum simulation, the time evolution is often precisely a parameter can be estimated.[3]
approximated through discrete time steps using the Trotter-
Suzuki decomposition: The Fisher information is intimately connected to the
expectation of the score operator ( 𝑆 ), which is defined as
𝑇/𝛿𝑡 the derivative of the logarithm of the likelihood function:
𝑒 −𝑖𝐻𝑡 ≈ (𝑒 −𝑖𝐻𝛿𝑡 )
1 𝑑
Where 𝑇 is the total evolution time and 𝛿𝑡 is the time 𝑆(𝜃) = √𝑝(𝜃)
√𝑝(𝜃) 𝑑𝜃
step. The Trotter-Suzuki approximation breaks down the
continuous evolution into a sequence of smaller steps,
simplifying the simulation process. However, long The Fisher information can be expressed as the
simulations can accumulate errors from each time step.[1] variance of the score operator:

The upper bound of the error introduced by the time 𝐹(𝜃) = Var(𝑆(𝜃))
block method can be estimated using the Lie-Trotter
formula, which provides an expression for the difference This relationship highlights the role of the Fisher
between the exact time evolution and the approximation: information in quantifying the information content of the
measurements with respect to the parameter 𝜃.
2 3 2𝑘−1
∥𝑒 −𝑖𝐻𝑡 − (𝑒 −𝑖𝐻𝛿𝑡 )𝑁blocks ∥ ≤ ∥ 𝐻 ∥ 𝑇 ( 𝑇 ) The Cramer-Rao bound provides a mathematical
∥ ∥ 3ℏ 𝑁blocks ℏ relation between the Fisher information and the achievable
precision of parameter estimation. For an unbiased estimator
Where ∥ 𝐻 ∥ is the operator norm of 𝐻 and 𝑘 is the 𝜃ˆ , the CramerRao bound states:
order of the Lie-Trotter formula. This derivation allows us
to quantify the accuracy of the time block method and 1
optimize the choice of time step, number of blocks, and Var(𝜃ˆ ) ≥
𝑁𝐹(𝜃)
order of the formula for a given simulation.

IJISRT24MAY1808 www.ijisrt.com 2190

Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1808

Where Var(𝜃ˆ ) is the variance of the estimator and 𝑁 is  The four-step procedure for information extraction
the number of measurements. serves as a fundamental framework in quantum
computation, encompassing the key stages of preparing
Quantum metrology techniques aim to approach the states, evolving quantum dynamics, measuring
CramerRao bound by optimizing measurement strategies observables, and optimizing parameters. This systematic
and exploiting quantum entanglement to enhance the Fisher approach underlies the advancements in variational
information. This enables quantum systems to achieve quantum algorithms and quantum metrology, driving the
measurements with unprecedented precision, surpassing development of accurate and precise quantum
classical limits. predictions.

 Four-Step Procedure for Information Extraction  Ramsey Interferometer Quantum Circuit in Experimental
The four-step procedure for information extraction in Setup
quantum computation involves a systematic approach to In experimental quantum metrology, the Ramsey
preparing quantum states, evolving parameterized quantum interferometer quantum circuit plays a pivotal role. The
states, measuring outputs, and estimating parameters based Ramsey interferometer is a fundamental quantum device
on multiple measurements. This procedure is integral to used to measure frequency and phase shifts with exceptional
variational parameter estimation and quantum metrology, precision [5]. It consists of two sequential applications of a
enabling the enhancement of predictive accuracy and 𝜋/2 pulse separated by a time delay 𝑇 and followed by a
precision in quantum computations. final 𝜋/2 pulse. This configuration effectively splits the
quantum state into two branches, allowing interference
 Preparation of Input States: The first step of the between the branches after the second pulse. By varying the
procedure involves the preparation of input quantum time delay 𝑇, the Ramsey interferometer becomes sensitive
states. These states serve as the initial conditions for the to small changes in frequency or phase.[5]
quantum computation. Parameterized quantum circuits Mathematically, the Ramsey interferometer can be
are used to generate these states, where the parameters 𝜃 represented as a sequence of unitary operators. Let 𝑈𝜋/2 be
determine the quantum state's characteristics (Fig.1. and the unitary operator corresponding to a 𝜋/2 pulse and 𝑈𝑇 be
Fig.2.). Variational techniques are applied to optimize the unitary operator corresponding to the time delay 𝑇. The
these parameters, ensuring that the prepared states are Ramsey interferometer circuit can be described as:
tailored to the specific problem at hand.
 Evolution of Parameterized Quantum States: Once the Ramsey Circuit = 𝑈𝜋/2 ⋅ 𝑈𝑇 ⋅ 𝑈𝜋/2
input states are prepared, the next step is to evolve them
over time using the Hamiltonian operator 𝐻. This time The unitary operators 𝑈𝜋/2 and 𝑈𝑇 can be represented
evolution is achieved through the application of quantum in matrix form, where 𝐻 is the Hadamard gate and 𝑒 −𝑖𝐻𝑡 is
gates that implement the unitary operator 𝑈(𝑡) = the time evolution operator with the Hamiltonian 𝐻 over
𝑒 −𝑖𝐻𝑡/ℏ . The parameterized nature of the quantum circuit time :
allows for flexibility in controlling the evolution
dynamics. The optimization of parameters using 𝜋
variational methods ensures that the quantum evolution 𝑈𝜋/2 = 𝐻 ⋅ 𝑒 −𝑖 4
approximates the desired transformation accurately. 𝑈𝑇 = 𝑒 −𝑖𝐻𝑇
 Measurement of Outputs: Following the evolution of
quantum states, measurements are performed to extract The interferometer's output state after the second 𝜋/2
relevant information. Observable quantities, represented pulse can be obtained by applying the Ramsey circuit to the
by Hermitian operators, are measured to obtain initial quantum state |𝜓⟩ :
measurement outcomes. These outcomes provide
insights into the quantum system's behavior and |𝜓out ⟩ = Ramsey Circuit ⋅ |𝜓⟩
dynamics. Quantum measurements introduce inherent
randomness due to the probabilistic nature of quantum The resulting state |𝜓out ⟩ exhibits oscillatory behavior
states, requiring multiple repetitions to gather sufficient as a function of the time delay 𝑇, allowing for the
statistical data.[4] measurement of phase shifts with high precision.
 Estimation of Parameters: The final step of the procedure
involves the estimation of parameters based on the Integrating the Ramsey interferometer quantum circuit
measurement outcomes. Estimators are used to infer the into the four-step procedure enhances the capabilities of
optimal parameter values that best align with the quantum metrology. The interferometer's sensitivity to
obtained measurements. Variational optimization phase shifts makes it a valuable tool for applications such as
techniques, such as gradient descent, are commonly atomic clocks, quantum sensors, and precision
employed to minimize the difference between the measurements in fundamental physics.
observed outcomes and the predicted outcomes from the
parameterized quantum circuit. This iterative process
refines the parameter estimates, leading to improved
accuracy and predictive power.

IJISRT24MAY1808 www.ijisrt.com 2191

Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1808

The four-step procedure, combined with the versatile

Ramsey interferometer, exemplifies the power of variational
quantum algorithms and quantum metrology. This approach
facilitates accurate predictions and measurements, with the
potential to revolutionize fields reliant on high-precision
data.

IV. GRAVITATIONAL WAVE MATCHED

FILTERING

We'll utilize the PyCBC library, a tool designed for

analyzing gravitational-wave data, identifying astrophysical
sources from compact binary mergers, and examining their
characteristics. These tools mirror those employed by the
LIGO and Virgo collaborations for detecting gravitational Fig 2 Waveform Corresponding Different Masses
waves within their data.[10]
Here, we've chosen to use the 'SEOBNRv4_opt' model.
We generate the waveform of a gravitational-wave Numerous alternatives are accessible, each employing
merger and matched filtering, which is optimal in the case of distinct approaches and encompassing varied physical
Gaussian noise and a known signal model. In reality our phenomena. This particular model simulates the
noise is not entirely Gaussian, and in practice we use a gravitational waveform produced by the merging of black
variety of techniques to separate signals from noise in holes, allowing for the spin of each black hole to align with
addition to the use of the matched filter. Here we generate the orbit. The specified parameters are: mass 1 = 20, mass
the gravitational waveform using one of the available 2 = 20, delta_ t = 1.0/4096, f_lower=40.. ( 𝐟𝐢𝐠 . 1)
waveform approximants. These can be generated as a time
series using get_td_waveform. The essential factors include It can be compared that the length of the waveform
the masses of the binary system (measured in solar masses), increases for lower mass binary mergers.(Fig. 2)
the sampling interval (in seconds), the initial frequency of
gravitational waves (in Hz ), and the choice of The distance of the waveform follows a simple linear
approximation method. Numerous approximation methods relation between distance (in megaparsec or Mpc ) and
are accessible, each accounting for various physical apmplitude when red-shift is not taken in consideration.(Fig.
phenomena. 3)

Then we reduced the dynamic range of the data and supress

low freqeuncy behavior which can introduce numerical
artefacts. We also downsampled the data to 2048 Hz as high
frequency content is not important.(Fig. 4)

Fig 3 Waveform Corresponding Different Distances

Fig 1 Waveform Generated

IJISRT24MAY1808 www.ijisrt.com 2192

Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1808

Fig 4 H1 Recorded GW150914 Merger Fig 6 Spectral Power Density

Observing a notable increase in the data at its edges, Conceptually, matched filtering involves laying the
we attribute this to the impact of the highpass and potential signal over your data and integrating (after
resampling processes employed in filtering. As the filter is weighting frequencies correctly). If there is a signal in the
applied to the edges, it loops back to the start of the data, data that aligns with the 'template', you will get a large value
resulting in a spike. This phenomenon occurs because the when integrated over. (Fig. 7)
data lacks cyclic continuity, causing the filter to resonate for
a duration equivalent to its length. While any visible Then SNR-signal to noise time series is calculated.
transients may not be apparent, it's imperative to avoid (Fig. 8) In our previous analysis, we identified a peak in the
filters operating on non-causally connected times. To signalto-noise ratio (SNR) concerning a suggested merger of
mitigate this issue, we opt to trim the data ends adequately binary black holes. Utilizing this SNR peak, we aim to align
to prevent wrapping around the input. This requirement will our proposal with the actual data and also to eliminate our
be consistently enforced across all filtering stages. (Fig. 5) proposal's influence from the data. To ensure a fair
comparison between the data and the signal, and to focus on
Effective matched filtering involves adjusting the the relevant frequency range, we intend to standardize both
weighting of frequency components in both the potential the template and the data through a process known as
signal and the data according to the noise amplitude. This whitening. Subsequently, we will filter both the data and the
process can be likened to filtering the data using a time template within the frequency range of 30 to 300 Hz. By
series version of the reciprocal of the Power Spectral doing so, any signal present in the data will undergo the
Density (PSD). To ensure control over the extent of this same transformation as the template, facilitating a more
filtering, we apply a windowing technique to the time equitable comparison. (Fig. 9)
domain equivalent of the PSD, limiting its length. While this
approach may result in some loss of information regarding
line behavior within the detector, the impact is minimal due
to the broad frequency range covered by our signals and the
narrowness of the lines.(Fig. 6)

Fig 7 Signal Model

Fig 5 Filter Wraparound

IJISRT24MAY1808 www.ijisrt.com 2193

Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1808

Fig 8 Signal to Noise Ratio

After alginment we subtract the template from original H1 data and results are obtained. (Fig. 10)

Fig 9 Data v/s Template

Fig 10 Original and Subtracted Signal

IJISRT24MAY1808 www.ijisrt.com 2194

Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1808

V. QUANTUM GRAVITATIONAL WAVE In order to understand the framework, we begin with

MATCHED FILTERING some essential preliminaries: N represents the number of
templates, while M signifies the number of data points in the
As previously explained, matched filtering involves time series. The data and templates are digitally encoded as
comparing time series data with templates to identify classical bits in the computational basis. Converting a
matches above a specified threshold. Gravitational wave classical, irreversible logic circuit into a reversible one is
(GW) data templates, which adhere to general relativity feasible, and this can be readily implemented on a quantum
principles, are computed on the fly rather than being computer by substituting classical reversible gates with their
preloaded from a database, as is often impractical due to quantum equivalents. This procedure demands some scratch
their sheer number. In a quantum context, this eliminates the space to facilitate reversible.
need for a massive data transfer into a quantum random-
access memory (qRAM). The process of constructing an
oracle to determine template matches is part of classical data
analysis and can be explicitly integrated without diminishing
the quantum approach's speedup potential. However, it's
crucial to note that the computational cost of an oracle call,
typically involving signal-to-noise ratio (SNR) calculations,
remains significant. Grover's algorithm doesn't expedite this
step. Still, quantum counting improves the computational
cost's dependence on the number of templates, making
previously unmanageable searches feasible. Importantly,
quantum counting [9] enables the detection of extremely
faint signals that classical methods can't discern, enhancing
the effectiveness of matched filtering in gravitational wave
data analysis.[11],[12],[13]
Fig 11 2 Qubits PQC Result-1
 Quantum Parameter Estimation
Estimators are used to infer the optimal parameter MSE vs 𝜙 plot for 2 Qubits Circuit,Learning Rate =
values that best align with the obtained measurements. 0.01
Variational optimization techniques, such as gradient
descent, are commonly employed to minimize the difference
between the observed outcomes and the predicted outcomes
from the parameterized quantum circuit. This iterative
process refines the parameter estimates, leading to improved
accuracy and predictive power.[4]

The need for careful LR tuning to ensure accurate

predictions and robust training in parameterized quantum
circuits is observed. Additionally, the quantum advantage
demonstrated by the parameterized quantum circuits,
showcases the potential of quantum computation techniques
in revolutionizing predictive modeling and achieving
unprecedented levels of accuracy and precision.

The 3-qubit quantum network consistently

outperformed the 2-qubit network with stable convergence, Fig 12 2 Qubits PQC Result-2
highlighting the benefits of increased qubit complexity in
capturing intricate data patterns. This study emphasizes the MSE vs 𝜙 plot for 2 Qubits Circuit,Learning Rate =
significance of tailored hyperparameters and model 0.04
architectures. While quantum networks offer intriguing
potential, meticulous tuning is crucial.  Calculations.
The implementation involves the use of four registers:
 Oracle Construction a data register (sized linearly with M ), an index register
To employ quantum counting in these applications, we (requiring log 2 N qubits), a template register (sized linearly
initiate the process with the creation of an oracle, which with M for intermediate calculations), and an SNR register
performs matched filtering while adhering to a predefined (with a constant size, O(1)).
threshold. The initial step involves outlining the pseudocode
for the construction of Grover's gate.[14]

IJISRT24MAY1808 www.ijisrt.com 2195

Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1808

The core of Grover's algorithm is the search over an  Step 0 (Initialization) [Cost: 𝑂(𝑀 + 𝑙𝑜𝑔 𝑁) ]:
index within a database, which necessitates the construction This step initializes the initial state, composed of four
of an oracle. The oracle calculates the template from an registers, and requires loading data, which takes linear time
index i, computes the SNR, and checks the result against a in M, and initializing the index register to an equal
predefined threshold. The number of gates required to superposition, which needs O(log N) gates.
compute a template waveform from its parameters is
denoted as k1, and it scales linearly with M. The calculation  Step 1 (Creating Templates) [Cost: 𝑂(𝑀) ]:
of the SNR between a template and the data requires k2 Calculating templates from the index is performed in
gates, with a time complexity of O(Mlog M). Finally, superposition over all index values, with a cost of k1
verifying whether the result surpasses a given threshold 𝜌 approximately O(M) gates. Step 2 (Comparison with the
thr, necessitating O(1) gates, is denoted as k3. The overall Data) [Cost: O(Mlog M) ): Calculating the SNR between the
computational complexity of the classical algorithm for template and the data costs k2 approximately O(Mlog M). It
computing matches against all templates is O(NMlog M). is then compared to a threshold to determine f(i), which
denotes whether a template is a match, with a cost of k3
approximately O(1).

 Step 3 (Disentangling Registers) [Cost: 𝑂(𝑀𝑙𝑜𝑔 𝑀) ]:

This step involves erasing intermediate calculations to
disentangle the index register from the other registers. The
erasure process has a cost of 𝑘1 + 𝑘2, which is roughly
𝑂(𝑀log 𝑀).

 Step 4 (Applying the Diffusion Operator) [Cost:

𝑂(𝑙𝑜𝑔 𝑁)] :
This step, unique to the quantum algorithm, requires
O(log N) quantum gates.

In total, the cost for a single oracle call is 𝑂(𝑀log 𝑀 +

log 𝑁). The integration of quantum counting in GW
matched filtering introduces efficiencies in match
determination and template retrieval. The construction of a
quantum oracle for this purpose enhances the computational
Fig. 13 3 Qubits PQC Result-1 process, ultimately improving the efficiency of the
algorithm.
MSE vs 𝜙 plot for 2 Qubits Circuit,Learning Rate =
0.01  Signal Detection
The primary focus is on signal detection, involving
four conditional probabilities.

 True Negative (P(r ∗ = 0 ∣ r = 0)) :

The probability of correctly determining that no
template exceeds the SNR threshold when there is no such
template.

 False Negative (P(r ∗ = 0 ∣ r > 0)) :

The probability of incorrectly identifying no match
when, in reality, no template exceeds the SNR threshold.

 True Positive (P(r ∗ > 0 ∣ r > 0)) :

The probability of correctly identifying templates
exceeding the SNR threshold when such templates exist.

 False Alarm (P(r ∗ > 0|r = 0)) :

The probability of incorrectly identifying templates
exceeding the SNR threshold when no such templates exist.

Fig 14 3 Qubits PQC Result-2 These probabilities differ from traditional definitions,
as they account for classification errors caused by the
MSE vs 𝜙 plot for 2 Qubits Circuit,Learning Rate = probabilistic nature of quantum algorithms. These
0.04 classification errors are our main focus in this study.

IJISRT24MAY1808 www.ijisrt.com 2196

Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1808

 Quantum Counting Procedure:  Procedure 2 (Retrieve All Matched Templates):

If the goal is to retrieve all matched templates, it's not
 Step 1: Initialization [Cost: 𝑂(𝑀 + log 𝑁)] - Setting the as straightforward as Procedure 1. It's akin to a "coupon
stage with data loading and index register initialization. collector problem," which requires approximately ( log 𝑟)
 Step 2: Creating Counting Register [Cost: O(1/ repetitions of Procedure 1. The complexity is similar for
2log N)] − Applying Hadamard gates to qubits, both procedures, provided the number of matching
incurring a cost of p. Step 3: Controlled Grover's Gate templates is significantly smaller than the total number of
[Cost: 𝑂((𝑀log 𝑀 + log 𝑁)/√𝑁)] - Determining the templates.
maximum iterations of Grover's gate required (2p − 1).
 Step 4: Inverse Quantum Fourier Transform [ Cost:  Probability of Failure:
𝑂((log 𝑁)2 )]. The section concludes by discussing the overall probability
 Step 5: Measurement [Cost: 𝑂(1/2log 𝑁)] - Calculating of failing to retrieve a matched template using this
b, the measurement outcome. procedure. It's emphasized that if this probability is less than
0.5 , then with a finite number of repetitions, it can be
reduced to a negligibly small value, ensuring successful
The overall cost of the algorithm is O(√𝑁(𝑀log 𝑀 +
retrieval of a matched template.
log 𝑁)).
 E. Grover's Algorithm and Matched Filtering
The choice of 'p' depends on the desired accuracy and
We try to find the frequency of a sine wave signal from
is related to the probability of false negatives (𝛿𝑛). A
amongst a number of known frequencies when the signal
specific value of 'p' is recommended to minimize false
start time and amplitudes are known. Time domain data D of
negatives while controlling computational cost.
length N is taken from the signal and compared to M
templates T of sine waves of known frequencies equally
The algorithm's cost is significantly lower than
spaced out in the frequency space of interest. The number of
classical approaches, making it efficient for signal detection
times Grover's algorithm is applied is dependent on P.
tasks involving a large number of templates (e.g., in
gravitational wave research). The computational cost
The signal sin wave looks like Fig 15.
difference is particularly prominent in cases with numerous
templates.[7],[11]
First let us make the state:
𝑀−1
 Matched Templates Retrieval 1
|𝜓ini ⟩ = ∑ |𝑚⟩|0⟩×𝑁
The process of retrieving matching templates relies on √𝑀 𝑚=0
Grover's algorithm from Algorithm 1 and utilizes the
estimated number of matching templates ('r*') from
This state is what the template information will be
Algorithm 2. While there are multiple approaches to retrieve
stored in. Let us also define:
matching templates when the number of matches is
unknown, this algorithm presumes that the signal detection 𝑀−1
algorithm is executed first, and its estimate of ' 𝑟 "' is 1
|𝑖⟩ = ∑ |𝑚⟩
naturally used for subsequent retrieval attempts. √𝑀 𝑚=0

 Algorithm 3 - Template Retrieval:

 Calculating the Number of Repetitions:

The initial step involves calculating the number of
repetitions required to retrieve the desired template (cost:
O(1) ). The value ' r ∗ ' ob − tained from Algorithm 2 is
used to determine the number of repetitions based on a
mathematical expression.

 Template Retrieval Complexity:

The algorithm's complexity is described as
O(√𝑁(𝑀log 𝑀 + log 𝑁)√𝑁 ), considering Algorithm 2 and
the retrieval of one template together.

 Procedure 1 (Retrieve One Template):

Repeat Grover's algorithm (Algorithm 1) ' k ∗ ' times Fig 15 Sin Signal
to obtain the desired template index (cost: O(√𝑁/𝑟 ∗
(𝑀log 𝑀 + log 𝑁)) ). The value of ' k ∗ ' is derived from This state has a basis state that corresponds to the
previous discussions and is proportional to √𝑁/r ∗. indexes of each template. We perform operation 𝑘ˆ1 on this
state to give a state that represent the template waveforms in
the frequency domanin |𝑇⟩ :

IJISRT24MAY1808 www.ijisrt.com 2197

Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1808

|𝑖⟩ ⊗ |𝑇⟩ = 𝑘ˆ1 (|𝑖⟩ ⊗ |𝜓ini ⟩) Assuming that every template is equally likely to have
the correct template without any more prior knowledge.
Similarly the data is loaded into a state represented by
|𝐷⟩. We get the template plot as shown in Fig. 16.  Create the Grover diffusion operator 𝑈ˆ𝑠 = 2|𝑠⟩⟨𝑠| − 𝐼.
 Apply 𝑈ˆ𝑤 then 𝑈ˆ𝑠 to |𝑠⟩𝑝 times to each state in 𝑃(𝑝 =
{0,1, … 𝑃 − 1}). Now we apply Grover's algorithm
itterably to this state such that:

|𝜓1 ⟩ = 𝐶ˆ𝐺 |𝜓0 ⟩

Where 𝐶ˆ𝐺 |𝑝⟩ ⊗ |𝑚⟩ → |𝑝⟩ ⊗ (𝐺ˆ )𝑝 |𝑚⟩

and 𝐺ˆ = 𝑈ˆ𝑠 𝑈ˆ𝑤

We can see what this operation does to |𝜓0 ⟩ in the plot

Fig. 17.

Fig 16 Templates

Grover's algorithm is applied after matched filtering is

to all templates. On a quantum computer, this can be done in
parallel to give an equal amplitude state |𝑤⟩ of length 𝑀 but
any state that corresponds to the index of a template that
meets the criteria of |𝑇 − 𝐷| < 10−3 has a phaseflip of -1 .
|𝑤⟩ is made by applying 𝑘ˆ2 :
𝑀
|𝑖⟩ ⊗ |𝑇⟩ ⊗ |𝐷⟩ ⊗ |𝑤⟩ = 𝑘ˆ2 (|𝑖⟩ ⊗ |𝑇⟩ ⊗ |𝐷⟩ ⊗ |0⟩×2 ).

Then we make the state: Fig 17 State Probability

𝑃−1 𝑀−1 We see the amplitudes of states 𝜓1 numbering 𝑃 × 𝑀.

1 Where the amplitudes of the states around correct solutions
|𝜓0 ⟩ = ∑ ∑ |𝑝⟩|𝑚⟩
√𝑃𝑀 𝑝=0 𝑚=0
are much greater than the incorrect solutions. The incorrect
solutions also exhibit a sinusoidal pattern.
This is equivalent of the state |𝑠⟩ previously made, but
𝑃 times. It is the frequency 𝑓 of this sinusoid that we wish to
determine, as it is related to the number of correct template
 The first part of Grover's algorithm is then applied as matches 𝑘 by:
follows:
𝑓𝜋
𝑘 = sin2
 Create operator 𝑈ˆ𝑤 = 𝐼 − 2|𝑤⟩⟨𝑤| where 𝑤 is the 𝑃
matrix position corresponding to the matching templates.
Determining phase/frequency information from
This operator has the property:
amplitudes of states requires a quantum fourier transform.
The quantum fourier transform is much the same as it's
𝑈ˆ𝑤 |𝑥⟩ = −|𝑥⟩ if 𝑥 = 𝑤
classical counterpart but is performed on amplitude/phase
𝑈ˆ𝑤 |𝑥⟩ = |𝑥⟩ if 𝑥 ≠ 𝑤 information stored on the states of qubits. It transfers
information stored in amplitudes in quantum states into
 Initiate superposition: phase information. There also exists the inverse quantum
𝑀−1
1 fourier transform for the reverse opperation. The quantum
|𝑠⟩ = ∑ |𝑥⟩ fourier transform acting on state |𝑝⟩ gives:
√𝑀 𝑥=0
𝐾−1
1 𝑘𝑥
QFT: |𝑥⟩ ↦ ∑ 𝑒 2𝜋𝑖 𝐾 |𝑘⟩
√𝐾 𝑘=0

IJISRT24MAY1808 www.ijisrt.com 2198

Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1808

The inverse quantum fourier transform is applied

across the ancillary qubits the recover the phase information Now we can just apply Grover's algorithm 𝑝opt times
from the sinusoidal behaviour in the states shown in the 1
to ∑𝑀−1 |𝑚⟩ to result in a state with amplitudes
graph above. This requires creating a 𝑄𝐹𝑇 −1 operator of √𝑀 𝑖=0
maximally amplified corresponding to matching states: We
size 𝑃 × 𝑃, which we will call 𝐹ˆ𝑃 as done below:
observe how the maximum state probability changes over
applications of Grover's iterations. This is plotted (Fig. 19)
Applying this across the ancillary states in |𝜓1 ⟩ :
along with a line indicating the determined optimal number
𝜓2 = 𝐹ˆ𝑃 𝜓1 of applications.
This operation changes |𝜓1 ⟩ to the states in the plot
(Fig. 18)

Fig 18 New State Probability Distribution Fig 19 State Probability over Grover Iterations

The probabilities |𝜓2 |2 are seen over ancillary and  F. Qiskit Implementation
template states. There are two peaks, corresponing to 𝑓 and Firstly, the registers are initialized. This is to set the
data we are matching against.
𝑷 − 𝑓. Measuring the ancillary state will likely acquire one
or the other. These peaks also correspond to the positions of
In the case of have multiple matching templates, we
matching templates, and so measuring the template qubits
shall not use 'search_Circuit.h(data[0])'. This will create two
likely recovers a correct position of a template.
data that we are matching against, rather than two matching
templates. In this case, both data would run as superpostions
From knowing 𝑓 or 𝑃 − 𝑓 we can work out the number
and so will the matching result. In essense, it is still one
of matched templates 𝑘 with the relation:
template searching so no untilization of quantum counting.
𝑓𝜋
𝑘 = 𝑀sin2 The Z-gate on the counting qubit is to compensate the
𝑃 general 𝜙 phase introduced by the Diffusion operator
function.
Doing this gives the number of matching templates.
Secondly, Grover's gate and Oracle function is
created.The oracle works by matching data and template bit
If the number is greater than 0 , we have matching
by bit in the qubit range within the precision. We use a C-X
templates.
gate for matching, in which case, the template qubits would
We find a corresponding matching template. To do this be, and only be in state 10⟩ if it is a match. After bit
we need to find out the optimum number of Grover's flipping, by applying a C-X gate ctrolled by all the template
applications to apply. This can easily be found from qubits, the ancilla qubit would be and only be flipped if all
knowing the number of matching templates: templates qubits matched. After the matching, we need to
reverse all the previous actions on the templates qubits.

𝜋 𝑀 Then, Diffusion operator function is created.

𝑝opt ≈ √
4 𝑘

IJISRT24MAY1808 www.ijisrt.com 2199

Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1808

Fig 20 Initializing Registers

Fig 21 Oracle Section 1

Fig 22 Oracle Section 2

IJISRT24MAY1808 www.ijisrt.com 2200

Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1808

Fig 23 Oracle Section 3

Fig 24 Oracle Section 4

Fig 25 Oracle Section 5

IJISRT24MAY1808 www.ijisrt.com 2201

Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1808

Fig 26 Oracle Section 6

Fig 27 Oracle Section 7

Fig 28 Oracle Section 8

IJISRT24MAY1808 www.ijisrt.com 2202

Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1808

Fig 29 Oracle Section 9

The diffusion operator can be written as: The measurement of the quantum counting process for
six-qubit data matching with a five-qubit counting register is
2|𝜓⟩⟨𝜓| − Î𝑁 = 𝐻ˆ ⊗𝑛 (2|0⟩⟨0| − Î𝑁 )𝐻ˆ ⊗𝑛 done. The first qubit is ignored to allow for two templates
matching.
Where |𝜓⟩ is the uniform superposition of states and 𝐼ˆ
is the N dimensional identity matrix. As 2|𝜓⟩⟨𝜓| − 𝐈ˆ𝑁 A measurement of the counting register in the
computational basis returns an integer value between 0 and
operates a reflection about the |𝜓⟩,2|0⟩⟨0| − 𝐈ˆ𝑁 ) operates a
2𝑝 − 1, from which we can now extract the desired estimate
reflection about the |0⟩. It turns out that Grover diffusion
of the phase. Intuitively, constructive interference occurs for
can be implemented on a quantum circuit with a phase shift
those elements {|𝑙′ ⟩} for which
operator that negates all the states except for |0⟩ sandwiched
between 𝐻ˆ ⊗𝑛 gates. 𝜃 𝑙′ 𝜋 − 𝜃 𝑙′
− 𝑝 ≃ 0 or − 𝑝≃0
𝜋 2 𝜋 2
However, this actually introduce an overall -1 to all
states. This would not change the results of Grover's We will only be interested in cases in which 𝑟 ≪ 𝑁,
searching part, but will affect the quantum counting part, and thus 𝜃 ≪ 1. Therefore, the observed measurement
introducing an overall 𝜋 phase. This is solved by applying a outcome, which we denote 𝑏, gives an unambiguous
Z-Gate to the lowest qubit in the counting register. (All the estimate of 𝜃, denoted 𝜃∗ as follows:
other qubits represents powers of two, which will result in
the Z-Gate being applyied even times, producing just a 1 𝑏𝜋
rather than -1 ) Third Step involves Quantum Counting 𝑝
, 𝑏 ⩽ 2𝑝−1
where we estimate the number of repetitions needed for 𝜃∗ = { 2
𝑏𝜋
multiple matches. It involves generation and application of 𝜋 − 𝑝 , 𝑏 > 2𝑝−1
controlled Grover's gate. We can use .to_gate() and 2
.control() to create a controlled gate from a circuit. We will
The theoretically most probable outcome b in this case,
call our Grover iterator grit and the controlled Grover
according to Equations above should be either 2 or 30 . The
iterator cgrit. All those controlled gate functions in qiskit is
most probable measurement result is 00010 , which in
difficult for this situation because they apply to gates, but
decimal is 2. The results are shown in Fig. 30 and Fig. 31.
our grovers operation is only a function. So either we can
rewrite this as a gate or we define our controlled funcion.
Then we define and apply QFT (Quantum Fourier
Transform) and measuring at last.

IJISRT24MAY1808 www.ijisrt.com 2203

Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1808

Fig 30 Measurement of Quantum Counting Process

Fig 31 Template Probabilities

IJISRT24MAY1808 www.ijisrt.com 2204

Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1808

Because there are two eigenvalues and we do not know which one is the measure value. We need to run them both and
choose the more reasonable one. Then we calculate and search for matched templates.

The measurement of Grover's search process for sixqubit data matching. The data is set as 000111 and the lowest qubit
is ignored to allow for two templates matching. With four iterations suggested by the quantum counting process as a numerical
output, the two templates that meet the matching criteria are returned with a probability higher than 99% altogether after 2048
trials on the ibmq_qasm_simulator(Fig. 32, Fig. 33).

Fig 32 Simulator Probabilities

Fig 33 Simulator Probabilities Enlarged

IJISRT24MAY1808 www.ijisrt.com 2205

 Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY1808

VI. CONCLUSION AND FUTURE PROSPECTS [10]. [10] J. Veitch, V. Raymond, B. Farr, W. Farr, P.
Graff, S. Vitale, B. Aylott, K. Blackburn, N.
The combination of quantum algorithms and quantum Christensen, M. Coughlin, W. Del Pozzo, F. Feroz, J.
metrology presents a promising avenue for advancing the Gair, C. J. Haster, V. Kalogera, T. Littenberg, I.
detection of gravitational waves. By utilizing Grover's Mandel, R. O'Shaughnessy, M. Pitkin, C. Rodriguez,
algorithm for signal detection and quantum metrology for C. Röver, T. Sidery, R. Smith, M. Van Der Sluys, A.
precise parameter estimation, researchers have demonstrated Vecchio, W. Vousden, and L. Wade, Parameter
significant improvements in sensor network sensitivity and estimation for compact binaries with ground-based
accuracy. This integrated approach effectively reduces gravitational-wave observations using the
noise, optimizes sensor placement, and enhances network LALInference software library, Phys. Rev. D 91,
scalability, laying a solid foundation for future gravitational- 042003 (2015).
wave astronomy research. Looking forward, the integration [11]. [11] G. Ashton, M. Hübner, P. D. Lasky, C. Talbot,
of advanced quantum techniques with gravitational-wave K. Ackley, S. Biscoveanu, Q. Chu, A. Divarkala, P.
detection holds immense potential for further progress. J. Easter, B. Goncharov et al., BILBY: A user-
Continuous refinement of quantum algorithms and friendly Bayesian inference library for gravitational-
metrology techniques is expected to lead to even greater wave astronomy, Astrophys. J. Suppl. 241, 27
improvements in signal detection efficiency and sensitivity. (2019).
Future investigations may explore additional quantum [12]. [12] T. Dal Canton and I. W. Harry, Designing a
algorithms, enhance the scalability of quantum sensor template bank to observe compact binary
networks, and apply these integrated methodologies to coalescences in Advanced LIGO's second observing
broader areas of observational astronomy. Moreover, run, arXiv:1705.01845 (2017).
advancements in quantum computing hardware will be [13]. [13] S. Dwyer, L. Barsotti, S. S. Y. Chua, M. Evans,
pivotal in unlocking the full capabilities of these techniques. [14]. M. Factourovich, D. Gustafson, T. Isogai, K.
As these technologies evolve, significant strides are Kawabe, A. Khalaidovski, P. K. Lam, M. Landry, N.
anticipated in expanding our comprehension of the universe Mavalvala, D. E. McClelland, G. D. Meadors, C. M.
and uncovering previously inaccessible cosmic phenomena. Mow-Lowry, R. Schnabel, R. M. S. Schofield, N.
Smith-Lefebvre, M. Stefszky, C. Vorvick, and D.
REFERENCES Sigg, "Squeezed quadrature fluctuations in a
gravitational wave detector using squeezed light,"
[1]. [1] Lloyd, S. (1996). Universal quantum simulators. Opt. Express 21, 19047-19060 (2013).
Science, 273(5278). [15]. [14] Chou, Chen-Kuan Chen, Wei Fwu, Peter Lin,
[2]. [2] Giovannetti, V., Lloyd, S., Maccone, L. (2004). Sung-Jan Lee, Hsuan-Shu Dong, Chen-Yuan. (2008).
Quantumenhanced measurements: Beating the Polarization ellipticity compensation in polarization
standard quantum limit. Science, 306(5700). second-harmonic generation microscopy without
[3]. [3] Paris, M. G., Rehacek, J. (2004). Quantum state specimen rotation. Journal of biomedical optics. 13.
estimation. Lecture Notes in Physics, 649. 014005. 10.1117/1.2824379.
[4]. [4] Childs, A. M., Kothari, R., Somma, R. D. (2018).
Quantum algorithm for systems of linear equations
with exponentially improved dependence on
precision. SIAM Journal on Computing, 47(5).
[5]. [5] Kaubruegger, R., Vasilyev, D. V., Schulte, M.,
Hammerer, K., Zoller, P. (2021). Quantum
variational optimization of Ramsey interferometry
and atomic clocks. Physical review X, 11(4), 041045.
[6]. [6] A. Montanaro, Quantum pattern matching fast on
average, Algorithmica 77, 16 (2017).
[7]. [7] B. P. Abbott, R. Abbott, T. D. Abbott, S.
Abraham, F. Acernese, K. Ackley, C. Adams, R. X.
Adhikari, V. B. Adya, C. Affeldt et al., Searches for
continuous gravitational waves from 15 supernova
remnants and Fomalhaut b with Advanced LIGO,
Astrophys. J. 875, 122 (2019).
[8]. [8] M. S. Anis, H. Abraham, R. A. AduOffei, G.
Agliardi, M. Aharoni, I. Y. Akhalwaya, G.
Aleksandrowicz, T. Alexander, M. Amy, S.
Anagolum et al., Qiskit: An open-source framework
for quantum computing, Zenodo Version: 0.7.2
(2019), doi:10.5281/zenodo.2562111.
[9]. [9] M. Mosca, Counting by quantum eigenvalue
estimation, Theor. Comput. Sci. 264, 139 (2001).

IJISRT24MAY1808 www.ijisrt.com 2206