Abstract
This paper introduces a quantum-secure scheme for conducting privacy-preserving maximum value determination, allowing the parties to ascertain the highest value from their confidential inputs while keeping non-maximum data private. Participants first transform their private inputs into binary representations and then apply local operations to encode these values into quantum states. These encoded states are transmitted to a semi-trusted intermediary, which computes the maximum through quantum–mechanical interactions. The protocol is designed to ensure confidentiality against potential external threats, including quantum attack strategies such as intercept-resend, entangle-measure, and Trojan horse attacks, while also preventing any disclosure of inputs between the participants. The framework is practical, utilizing entangled Bell states as carriers of information, leveraging quantum state manipulations for secure encoding, and employing quantum measurements for result extraction. Empirical results obtained from simulations using IBM Quantum Composer demonstrate the operational feasibility of the scheme. A correctness analysis indicates that if the participants accurately submit their respective inputs and adhere strictly to the protocol, the semi-trusted intermediary will reliably reveal the maximum value to both parties. Additionally, fairness is ensured through the intermediary’s role in disclosing the maximum value simultaneously to both parties, preventing any unfair advantage.
1 Introduction
Recent advancements in quantum communication [1, 2] and computation [3, 4] have catalyzed transformative developments in cryptography and data processing. Quantum cryptography [5, 6], for instance, exploits quantum mechanical principles to achieve unprecedented security in information exchange, while quantum machine learning [7, 8] merges quantum computational power with classical algorithms to enhance analytical capabilities.
However, the rapid evolution of quantum computing jeopardizes classical cryptographic frameworks reliant on mathematical hardness assumptions, such as RSA and elliptic curve cryptography. Shor’s algorithm [9], capable of factoring large integers exponentially faster than classical methods, directly threatens public-key encryption. Similarly, Grover’s algorithm [10] diminishes the effective security of symmetric-key systems by accelerating unstructured search tasks.
Quantum cryptography emerges as a robust countermeasure, offering protocols like BB84 [11] that ensure information–theoretic security [12] by detecting eavesdropping through quantum-state disturbances. This capability stems from fundamental principles like the no-cloning theorem and wavefunction collapse, which prevents undetected interception of quantum transmissions. Beyond secure communication, quantum cryptography enables functionalities such as quantum key distribution (QKD) [13, 14], quantum secure direct communication (QSDC) [15, 16], quantum key agreement (QKA) [17, 18], quantum private comparison (QPC) [19–27], and quantum private set intersection (QPSI) [28].
Secure multiparty computation (MPC) represents a critical area within modern cryptography, enabling multiple distrustful parties to jointly compute a function without disclosing their private inputs [29]. MPC has applications in various fields, including privacy-preserving data analysis [30–32], secure voting systems [33, 34], and deep learning [35–37]. This concept of MPC was pioneered by Andrew Yao through his foundational work on the Millionaires’ problem [38], where two parties aim to determine who possesses greater wealth without revealing their actual financial amounts. Yao’s formulation established the theoretical basis for numerous MPC protocols, illustrating how collaborative computation can be achieved while preserving data confidentiality. A natural extension of this problem involves privately determining the maximum value among multiple inputs, where participants seek to identify the largest value without exposing their individual data. Classical approaches to this task rely on private comparison protocols [39–43], which compute the maximum value using traditional cryptographic methods. However, these methods depend on unproven computational assumptions, rendering them susceptible to quantum-based attacks. The advent of quantum computing, with its capacity to efficiently solve problems like integer factorization and discrete logarithms, poses a significant threat to the security of classical cryptographic systems.
The computation of maximum values holds significant importance in privacy-sensitive applications, such as sealed-bid auctions [44], electronic voting [45], and federated learning [46]. Although quantum algorithms [47, 48] have been developed to efficiently identify maximum and minimum values within a dataset, they lack mechanisms to protect the privacy of individual inputs during computation. To date, no established quantum scheme that specifically addresses the privacy-preserving computation of the maximum value between two private inputs exists.
To bridge this gap, this paper introduces a novel quantum-based scheme designed to determine the maximum value while safeguarding the privacy of non-maximum inputs. By leveraging the principles of quantum mechanics, the proposed scheme allows two users to collaboratively compute the maximum value without disclosing their private data. Security analyses confirm that the scheme is resilient to both external and internal attacks. Furthermore, the scheme employs single-photon operators, including Pauli operators and the Ry rotation operator, alongside Bell states as quantum resources. It utilizes single-photon operations and Bell measurements, making it feasible to implement with current quantum technologies. The practicality of the scheme is validated through simulations conducted on the IBM Quantum Cloud Platform.
The structure of this paper is as follows: Section 2 provides the necessary background on unitary operations and encoding techniques. Sections 3 and 4 detail the proposed scheme and its simulation, respectively. Section 5 presents an analysis of the scheme’s correctness, security, fairness, and qubit efficiency. Section 6 provides the conclusions of the paper.
2 Preliminaries
The Ry rotation operator around the y-axis in the Bloch sphere [49] can be represented by
This operator can be considered an encryption operator used to convert a quantum state into an unknown quantum state, and the encryption key is . The Ry rotation operator applied to the unknown quantum states can recover the original quantum states. The Ry rotation operator is symmetric with respect to the Z-axis, ensuring balanced access to the qubit states. This symmetry is instrumental in minimizing bias toward specific measurement outcomes, which is essential for maintaining security. Furthermore, users can encrypt qubits using the Ry rotation operator, complicating the eavesdropper’s (often referred to as “Eve”) ability to predict results without inducing measurement disturbances. When Eve attempts to measure the qubits, she inevitably alters their states due to the no-cloning theorem and the disturbances associated with measurement. Consequently, the Ry rotation operator guarantees that any measurement attempt by Eve will result in an increased error rate, which can be detected by users, thereby enhancing the security of quantum communication protocols.
Three operations are as follows:
•I Gate: . This gate leaves the quantum state unchanged.
•X Gate: . This gate flips the states, converting to and vice versa.
•Z Gate: . This gate changes to and to .
The three local operations written in a matrix form are as follows:
Four Bell states are two-qubit entangled states, which can be written as
We suppose that a third party (TP) prepares a Bell state in and distributes the first qubit to Alice and second qubit to Bob, where Alice processes her own bits and Bob has his own bit . There are three encoding rules [50] corresponding to Alice and Bob’s bits, as shown in Table 1. When Alice and Bob apply one of the three encoding rules to , we can see that if and only if and can make the state, changes to . This is the key to design the following scheme.
TABLE 1
| (1) | (2) | (3) | ||||||
Three different encodings.
3 The proposed scheme
Consider a scenario involving two parties, Alice and Bob, each possessing private inputs A and B, respectively, where the upper bound of these values is N. Their objective is to determine the maximum value between A and B with the assistance of a semi-honest TP, such as a quantum server, while ensuring the privacy of non-maximum inputs. The proposed privacy-preserving maximum value determination scheme must satisfy the following requirements:
Correctness: If Alice and Bob honestly submit their respective inputs A and B and adhere to the protocol, the semi-honest TP will accurately disclose the maximum value to both parties.
Fairness: The protocol must ensure that no party gains an unfair advantage over the other. Each participant should receive the computation outcome without concerns of manipulation or bias.
Privacy: Except for the maximum value, no party should gain any information about the non-maximum inputs, even in the presence of internal or external eavesdroppers.
The scheme operates under the honest-but-curious model, where participants follow the protocol but may attempt to infer private information from the execution. The semi-honest TP is equipped with quantum capabilities and is responsible for preparing quantum resources, such as generating Bell states, and performing Bell measurements. Alice and Bob, also possessing quantum capabilities, perform basic single-photon operations and prepare decoy photons for eavesdropping detection. We assume that the protocol operates over an ideal channel with no noise [51]. However, in practical scenarios, quantum error-correcting codes [52, 53] can be implemented to detect and correct errors induced by noise. Additionally, Alice and Bob share a secret key , where each for . The steps of the proposed scheme are as follows, and its diagram is illustrated in Figure 1.
FIGURE 1
The diagram of the scheme.
Step 1Alice and Bob convert their private inputs A and B into N-bit strings and , respectively. For each input, positions corresponding to the input value and all positions to its right are marked as 1, while positions to the left are marked as 0. For example, if and with an upper bound of 6, the resulting bit strings are and , respectively.
Step 2TP generates N Bell states in and distributes the first and second particles of these states into two quantum sequences, and . To detect eavesdropping, TP prepares decoy particles, randomly chosen from the set . TP records the initial states of these decoy particles and inserts decoy particles into , while another decoy particles are inserted into at random positions, resulting in new sequences and . TP then sends and to Alice and Bob, respectively.
Step 3Upon receiving (), Alice (Bob) interacts with the TP for conducting the eavesdropping detection. TP announces the inserted positions and the measurement basis of the decoy particles to Alice (Bob). If the decoy particle is chosen in states , the measurement basis is Z-basis. Otherwise, the measurement basis is X-basis. Alice (Bob) conducts the corresponding quantum measurements on the decoy particles and returns the measurement results back to the TP who calculate the error rate. The error rate is defined as the fraction of bits incorrectly received by Alice (Bob) compared to what the TP sent. In the presence of eavesdropping, this error rate increases due to the additional disturbances introduced by the measurements of eavesdroppers (often termed “Eve”). If the error rate exceeds a certain threshold, it indicates that eavesdropping exists, and the TP tells Alice (Bob) to abort and restart the scheme from Step 2. Otherwise, the scheme goes to Step 4.
Step 4
Alice (Bob) recovers
(
by discarding the
decoy particles inserted in
(
) and performs the following operations:
(1) Alice (Bob) chooses one of the three encoding rules in Table 1 based on the shared secret key to encode her (his) N-bit strings (.
•If , the first encoding rule is chosen.
•If , the second encoding rule is chosen.
•If , the third encoding rule is chosen.
(2) Alice applies the unitary operations corresponding to the chosen encoding rule in (1) to the received sequence to generate the encoded sequence . Similarly, Bob applies the unitary operations corresponding to the chosen encoding rule in (1) to the received sequence to generate the encoded sequence .
(3) Alice prepares a secret key , where for , and applies the Ry rotation operators to , where can be considered an encryption key for encrypting . The new generated sequence is denoted as . Similarly, Bob prepares a secret key , where for , and applies the Ry rotation operators to , where can be considered an encryption key for encrypting . The new generated sequence is denoted as .
(4) Alice (Bob) prepares decoy particles, each of which is randomly chosen from , records the initial decoy states of these decoy particles, and inserts these particles into ( in a random position. The new generated sequence is denoted as .
(5) Alice (Bob) records the inserted positions of these decoy particles and sends to the TP.
Step 5After receiving , the TP interacts with Alice (Bob) for conducting eavesdropping detection in the same manner in step 3. If no eavesdropper is present during the transmission of the quantum sequence, Alice shares and Bob shares to the TP for further processing in the next step. Otherwise, the scheme is aborted and restarted from Step 2.
Step 6
TP recovers
(
by discarding the
decoy particles inserted in
and performs the following operations:
(1) TP applies the Ry rotation operators on and on to recover and , respectively. This process can be considered a decryption of and .
(2) TP conducts Bell measurements on and to obtain the measurement results, which consist of N Bell states.
(3) If the first occurrence of the j-th position of the measurement result is , the TP identifies j as the maximum value of the private inputs from Alice and Bob.
(4) TP announces the maximum value to Alice and Bob.
4 Simulation
Considering a case that Alice and Bob have their private inputs A = 3 and B = 1, respectively, the upper bound of the values A and B is 3 (since the available qubits in IBM Quantum Composer are seven). They intend to determine the maximum value of their inputs with the assistance of a TP while ensuring the privacy of non-maximum inputs. We assume that the secret key shared between Alice and Bob via a QKD protocol is .
As described in our scheme, Alice and Bob convert their private inputs A and B into two 3-bit strings and . The encoding rules based on the shared secret key selected by Alice and Bob is the second, third, and first encoding in Table 1. Therefore, we can determine that the corresponding unitary operations performed on the received sequences and are {I, X, X} and {X, X, Z}, respectively. We assume that the secret keys prepared by Alice and Bob are and , respectively. The measurement results conducted by Bell measurements yield 00, 01, 10, and 11 corresponding to , and , respectively.
We simulate this scenario using IBM Quantum Composer, concentrating on the quantum operations while excluding eavesdropping detection, as this is a separate process within quantum communication protocols. The quantum circuit designed to determine the maximum value between inputs A and B is illustrated in Figure 2. Additionally, Figure 3 presents a histogram displaying the probability of measuring each state.
FIGURE 2
![Quantum circuit diagram with six qubits labeled q[0] to q[5]. It features Hadamard (H) gates, controlled gates, and Ry rotation gates with different angles. Measurements are shown in gray blocks.](https://www.frontiersin.org/files/Articles/1592890/xml-images/fphy-13-1592890-g002.webp)
The quantum circuit simulation.
FIGURE 3
The probability of measuring each state.
According to the results depicted in Figure 3, the final measurement results, from right to left, are 10, 00, and 11, corresponding to , and , respectively. Therefore, we can determine that the first occurrence of the measurement result in the third position is , leading us to deduce that the maximum value of two private inputs from Alice and Bob is 3.
5 Analysis
5.1 Correctness
Alice and Bob each have private inputs A and B. They convert these inputs into two N-bit strings and . The maximum value j is defined as the starting position of the first occurrence, such that both and have “1.”
For instance, we assume that the upper bound of the values A and B is 6, where and . The two 6-bit strings and are and , respectively. In this case, has “1” s at positions 3, 4, 5, and 6. has “1” s at positions 5 and 6. We know that the starting position of the first occurrence such that both and have “1” is 5. Therefore, the maximum is 5.
When designing the privacy-preserving maximum value determination scheme, we only need to find the starting position of the first occurrence such that both and have “1.” According to the three encoding rules corresponding to Alice and Bob’s bits, we know that if the j-th position of both and have “1,” the initial state changes to . The proposed scheme is designed based on the above correctness analysis, while the Ry rotation operators and decoy particles are used for ensuring privacy of the private inputs.
Therefore, our scheme is correct, provided that Alice and Bob honestly submit their respective sets A and B and adhere to the protocol. Ultimately, the semi-honest third party will disclose the maximum value to both Alice and Bob.
5.2 Security
In our scheme, except for the maximum value, no party learns any information about the other non-maximum inputs, even when facing threats from anyone within the group of participants or from outside eavesdroppers. Therefore, the proposed scheme must satisfy the following security requirements:
(1) Any attempt by an eavesdropper to gain information about the private inputs would be detectable, even if the eavesdropper possesses quantum capabilities and can perform various quantum-attack strategies.
(2) The semi-honest third party has no viable means of obtaining the private inputs, except for the maximum value.
(3) Alice cannot access Bob’s private inputs unless the maximum value corresponds to Bob’s private inputs.
(4) Bob cannot access Alice’s private inputs unless the maximum value determined by the protocol matches Alice’s private inputs.
Any attempt by an eavesdropper to gain information about the private inputs would be failed, even if the eavesdropper possesses quantum capabilities and perform various quantum-attack strategies.
Proof. In quantum communication, the presence of an outsider eavesdropper, commonly referred to as Eve, poses a significant risk. Eve may apply various quantum-based attacks, including intercept-resend [54, 55], direct-measure [56], entangle-measure [57] and quantum Trojan horse attacks [58], to attempt obtaining the private inputs. To counter these eavesdropping strategies, the participants use the decoy-state method. This technique involves sending additional decoy states along with the actual quantum states. Because of the decoy-state method, Eve’s various quantum-attack strategies (intercept-resend, measure-resend, direct-measure, and entangle-measurement) are rendered ineffective. Alice and Bob can reliably detect any eavesdropping attempts through discrepancies in the statistical properties of the decoy states. This ensures security of their communication.
5.2.1 Case I: the intercept-resend attack
In this scenario, Eve intercepts the quantum particles sent from TP to the participants. She stores these intercepted particles for future use. After storing the intercepted quantum states, Eve prepares her own single particles, which she sends to the participants in place of the original particles. After the participants perform their operations on the particles sent by Eve, Eve intercepts the quantum particles sent back to the TP and measures these particles to extract secret information, while returning her previously stored particles to the TP. Upon receiving the quantum particles, the participants and TP initiate eavesdropping detection. The TP discloses the positions and measurement bases of the decoy particles used during communication. Since Eve does not know the specific states of these particles, there is a 50% probability that the participants will obtain an incorrect result when measuring the particles sent by Eve. For example, if the original decoy photon is in the state ∣1⟩, but Eve prepares a particle in the state ∣+⟩, the measurement performed by the participants on Eve’s particle has a 50% chance of yielding an incorrect result. The probability that Eve can pass the detection is given bywhere is the number of decoy photons consumed. The relationship between the number of decoy photons and the probability of Eve successfully deceiving the detection is shown in Figure 4.
FIGURE 4
The relationship between the number of decoy photons and the probability of Eve successfully deceiving the detection.
When , . As becomes large enough, the probability that Eve can pass the detection approaches 0. Therefore, this attack is invalid under this communication method, ensuring security of the participants’ private information.
5.2.2 Case II: the direct-measure attack
After performing the intercept-resend attack to obtain the positions of the decoy particles, Eve can discard the decoy particles in and to obtain the encoded quantum sequence and , respectively. Although this attack has been detected, Eve has obtained and and may attempt a direct-measure attack on and . However, quantum measurements collapse quantum states, and the measurement outcomes are intrinsically linked to the secret keys chosen by Alice and Bob. These secret keys will remain undisclosed if the eavesdropping detection is not passed. Due to the challenges of distinguishing decoy photons, the eavesdropping detection mechanism, and the reliance on secret keys for confidentiality, Eve’s direct-measurement attack is ultimately ineffective.
5.2.3 Case III: the entangle-measure attack
Eve may also attempt an entangle-measure attack by entangling her auxiliary quantum particles with the intercepted particles. This allows her to extract information by measuring her auxiliary particles. In our scheme, a bidirectional quantum channel is used for transmitting quantum sequences, enabling Eve’s entangle-measure attack to be modeled using two unitary operations, and Here, is applied to the channel from the TP to Alice and Bob, while is applied to the channel from Alice and Bob back to the TP. As an example, consider that is applied to the channel from the TP to Alice and Bob. When Eve uses to entangle her particles with the four possible states , the resulting states are as follows:
The coefficients must satisfy normalization conditions: and . For Eve’s attack to go undetected, the following conditions must be satisfied.
Substituting the results in Equation 5 into Equations 1–4, the results are given in the following equations:
According to Equations 6–9, Eve’s auxiliary particles remain independent from the target particles For Eve to avoid introducing errors during eavesdropping detection. This lack of entanglement indicates that there is no quantum correlation between Eve’s measurements and the target particles, and Eve cannot gain information about the target particles by measuring her auxiliary particles. As a result of this independence and lack of entanglement, Eve’s attempts to conduct an entanglement attack are ineffective.
5.2.4 Case IV: the Trojan horse attacks
Trojan horse attacks, including the delay-photon attack and invisible photon attack [59], mainly occur in the bidirectional quantum channel that is used for transmission of quantum sequence. Since the quantum sequences transmitted in our scheme are TP–Alice/Bob–TP, our scheme is vulnerable to these attacks. The implementation of a wavelength quantum filter and a photon number splitter [60] significantly enhances the security of the two-way quantum protocol against Trojan horse attacks. By ensuring that only legitimate quantum states are transmitted, the protocol maintains its privacy.
Therefore, any attempt by an eavesdropper to gain information about the private inputs would fail, even if the eavesdropper possesses quantum capabilities and can perform various quantum-attack strategies.
The semi-honest third party has no viable means of obtaining the private inputs, except for the maximum value.
Proof. In the proposed quantum scheme, a semi-honest TP might attempt to gain information about Alice’s and Bob’s private inputs by preparing single photons instead of the intended Bell states. However, the protocol incorporates mechanisms that ensure security even in the presence of such attacks. In our scheme, TP prepares 2N single photons, distributing N photons to Alice and N to Bob. This setup allows the TP to potentially manipulate the quantum states being transmitted. When Alice and Bob apply their respective unitary operations, based on the secret key K, to encode their inputs into states and , the TP can obtain the resultant states post-encoding. Unfortunately, different unitary operations performed on the same single photons can yield the same measurement results, as shown in Table 2. For instance, the unitary operations I and Z performed on state result in the final state also being . As a result, the proposed scheme remains secure against attacks from the TP. The necessity of the secret key K for proper encoding ensures that the TP does not gain any meaningful information about Alice’s and Bob’s private inputs, except for the maximum value.
TABLE 2
| (1) | (2) | (3) | ||||||
The resultant states.
Alice cannot access Bob’s private inputs unless the maximum value corresponds to Bob’s private inputs.
Proof. In the proposed quantum scheme, Alice may attempt an intercept-resend attack to gain access to Bob’s private inputs by intercepting transmitted from Bob to the TP. To execute the attack, Alice intercepts the quantum sequence and replaces it with a fake sequence before sending it to TP. Bob, upon receiving the potentially altered sequence, will still announce the positions of the decoy particles used for eavesdropping detection. However, the scheme is designed to detect any eavesdropping. If Bob identifies discrepancies during the eavesdropping detection phase, he will recognize that an attack has occurred. If Alice somehow manages to ascertain from the sequence , she may attempt to deduce Bob’s private inputs using the relationship between and the shared secret key K. However, this is not straightforward. Since Bob encrypts his input with the secret key using the Ry rotation operators before any communication takes place, this encryption ensures that even if Alice intercepts the data, she cannot decrypt or understand it without the appropriate key. The condition for Bob to announce the secret key is that the eavesdropping detection must pass. If the detection phase fails and an eavesdropper is present, Bob will not disclose his secret key . Alice’s intercept-resend attacks position her as an eavesdropper; therefore, the secret key will not be announced.
Therefore, Bob’s private inputs are inaccessible to Alice unless the maximum value corresponds to Bob’s private inputs.
Bob cannot access Alice’s private inputs unless the maximum value determined by the protocol matches Alice’s private inputs.
Proof. In the proposed quantum scheme, both Alice and Bob have identical roles. If Bob attempts to access private inputs, he may try to execute an intercept-resend attack by intercepting the quantum sequence that Alice transmits to TP. Just as with Alice’s attack, although Bob can obtain the sequence by performing the intercept-resend attack, he cannot succeed to obtain . For Bob to successfully decrypt to retrieve , he would need to know Alice’s secret key . If the eavesdropping detection is passed, Bob will be able to access Alice’s secret key because it is only disclosed after ensuring that the communication is secure. However, Bob fails to pass this detection due to the execution of the intercept-resend attack, and as a result, the secret key will not be announced. Therefore, Alice’s private inputs are inaccessible to Bob.
5.3 Fairness
Fairness is a critical aspect of any secure computation protocol. In the proposed scheme, the involvement of the third party plays a pivotal role in ensuring that both Alice and Bob have equal access to the maximum value. After measuring the quantum sequence, the third party publishes the results and ensures that both Alice and Bob receive this information simultaneously, thereby preventing any one party from gaining an undue advantage over the other. Consequently, fairness in the proposed scheme is assured through the third party’s role in simultaneously publishing the maximum value for both participants.
5.4 Qubit efficiency
The qubit efficiency, an important indicator for measuring the utilization rate of qubits, can be defined bywhere q represents the qubit efficiency, c denotes the upper bound of the input values, and t denotes the number of qubits consumed during the process of determining the maximum value, excluding those designated for eavesdropping detection during the transmission of quantum sequence. In our scheme, N Bell states all in as quantum resources, enabling the determination of the maximum with an upper bound of N; thus, we can derive c = N, t = 2N. Consequently, the qubit efficiency of the proposed scheme is .
6 Conclusion
In this work, we introduce a privacy-preserving scheme for determining the maximum value between private inputs while safeguarding the confidentiality of non-maximum values. The proposed scheme employs Bell states as quantum resources, Pauli operators for encoding inputs, the Ry rotation operator for encryption, and Bell measurements to extract the results. This design ensures compatibility with existing quantum technologies, and its practical feasibility has been validated through simulations conducted on the IBM Quantum Cloud Platform. A correctness analysis confirms that if Alice and Bob faithfully provide their respective inputs A and B and strictly follow the protocol, the semi-honest TP will accurately reveal the maximum value to both parties. Security analysis further demonstrates that, beyond the maximum value, no participant or external eavesdropper can gain access to information about the non-maximum inputs, even under adversarial conditions. Fairness is guaranteed by the TP’s role in simultaneously disclosing the maximum value to Alice and Bob, ensuring that neither party gains an unfair advantage. However, it is worth noting that the scheme assumes honest participation. If one party acts rationally, for example, by submitting a false input, the integrity of the other party’s input could be compromised. To address this limitation, future research will explore quantum schemes involving rational participants who may exhibit selfish behavior, aiming to enhance robustness and fairness in such scenarios.
Statements
Data availability statement
The original contributions presented in the study are included in the article/supplementary material; further inquiries can be directed to the corresponding author.
Author contributions
MH: Methodology, Formal Analysis, Writing – original draft, Investigation, Conceptualization, and Supervision. YW: Funding acquisition, Writing – review and editing.
Funding
The author(s) declare that financial support was received for the research and/or publication of this article. This research is supported by the Open Fund of Network and Data Security Key Laboratory of Sichuan Province (Grant No. NDS2024-1) and the Gongga Plan for the “Double World-class Project.”
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declare that no Generative AI was used in the creation of this manuscript.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
1.
ChenYAZhangQChenTYCaiWQLiaoSKZhangJet alAn integrated space-to-ground quantum communication network over 4,600 kilometres. Nature (2021) 589(7841):214–9. 10.1038/s41586-020-03093-8
2.
PanJWSimonCBruknerČZeilingerA. Entanglement purification for quantum communication. Nature (2001) 410(6832):1067–70. 10.1038/35074041
3.
GrahamTMSongYScottJPooleCPhuttitarnLJooyaKet alMulti-qubit entanglement and algorithms on a neutral-atom quantum computer. Nature (2022) 604(7906):457–62. 10.1038/s41586-022-04603-6
4.
DaleyAJBlochIKokailCFlanniganSPearsonNTroyerMet alPractical quantum advantage in quantum simulation. Nature (2022) 607(7920):667–76. 10.1038/s41586-022-04940-6
5.
PortmannCRennerR. Security in quantum cryptography. Rev Mod Phys (2022) 94(2):025008. 10.1103/revmodphys.94.025008
6.
MehicMMichalekLDervisevicEBurdiakPPlakalovicMRozhonJet alQuantum cryptography in 5G networks: a comprehensive overview. IEEE Commun Surv & Tutorials (2023) 26:302–46. 10.1109/comst.2023.3309051
7.
HuangXZhangSLinCXiaJ. Quantum fuzzy support vector machine for binary classification. Comput Syst Sci Eng (2023) 45(3):2783–94. 10.32604/csse.2023.032190
8.
CerezoMVerdonGHuangHYCincioLColesPJ. Challenges and opportunities in quantum machine learning. Nat Comput Sci (2022) 2(9):567–76. 10.1038/s43588-022-00311-3
9.
ShorPW. Polynomial-time algorithms for prime factorization and discrete logarithms on a quantum computer. SIAM Rev (1999) 41(2):303–32. 10.1137/s0036144598347011
10.
GroverLK. Quantum mechanics helps in searching for a needle in a haystack. Phys Rev Lett (1997) 79(2):325–8. 10.1103/physrevlett.79.325
11.
BennettCHBrassardG. Quantum cryptography: public key distribution and coin tossing. Theor Comput Sci (2014) 560:7–11. 10.1016/j.tcs.2014.05.025
12.
LuoYLiQMaoHK. Distributed information-theoretical secure protocols for quantum key distribution networks against malicious nodes. J Opt Commun Networking (2024) 16(10):956–68. 10.1364/jocn.530575
13.
ZhangWvan LeentTRedekerKGarthoffRSchwonnekRFertigFet alA device-independent quantum key distribution system for distant users. Nature (2022) 607(7920):687–91. 10.1038/s41586-022-04891-y
14.
NadlingerDPDrmotaPNicholBCAranedaGMainDSrinivasRet alExperimental quantum key distribution certified by Bell's theorem. Nature (2022) 607(7920):682–6. 10.1038/s41586-022-04941-5
15.
HuangXZhangSChangYYangFHouMChengW. Quantum secure direct communication based on quantum homomorphic encryption. Mod Phys Lett A (2021) 36(37):2150263. 10.1142/s0217732321502631
16.
ShengYBZhouLLongGL. One-step quantum secure direct communication. Sci Bull (2022) 67(4):367–74. 10.1016/j.scib.2021.11.002
17.
HuangXZhangSBChangYQiuCLiuDMHouM. Quantum key agreement protocol based on quantum search algorithm. Int J Theor Phys (2021) 60:838–47. 10.1007/s10773-020-04703-x
18.
LinSZhangXGuoGDWangLLLiuXF. Multiparty quantum key agreement. Phys Rev A (2021) 104(4):042421. 10.1103/physreva.104.042421
19.
WangBLiuSQGongLH. Semi-quantum private comparison protocol of size relation with d-dimensional GHZ states. Chin Phys B (2022) 31(1):010302. 10.1088/1674-1056/ac1413
20.
GongLHLiMLCaoHWangB. Novel semi-quantum private comparison protocol with Bell states. Laser Phys Lett (2024) 21(5):055209. 10.1088/1612-202x/ad3a54
21.
WangBGongLHLiuSQ. Multi-party semi-quantum private comparison protocol of size relation based on two-dimensional Bell states. Chin Phys B (2024) 33(11):110303. 10.1088/1674-1056/ad73ae
22.
ZhouNRChenZYLiuYYGongL. Multi‐party semi‐quantum private comparison protocol of size relation with d‐level GHZ states. Adv Quan Tech (2024) 8:2400530. 10.1002/qute.202400530
23.
HuangXZhangSBChangYHouMChengW. Efficient quantum private comparison based on entanglement swapping of bell states. Int J Theor Phys (2021) 60:3783–96. 10.1007/s10773-021-04915-9
24.
HouMWuY. Single-photon-based quantum secure protocol for the socialist millionaires’ problem. Front Phys (2024) 12:1364140. 10.3389/fphy.2024.1364140
25.
HuangXZhangSBChengW. Quantum private comparison based on GHZ-type states[C]//2021 IEEE AFRICON. IEEE (2021). p. 1–4.
26.
HouMWuYZhangS. New quantum private comparison using four-particle cluster state. Entropy (2024) 26(6):512. 10.3390/e26060512
27.
LianJYYeTYYeCQ. Circular semiquantum private comparison protocol for equality without a preshared key based on χ-type states. Phys Rev Appl (2025) 23(4):044006. 10.1103/physrevapplied.23.044006
28.
HuangXZhangWZhangS. Quantum multi-party private set intersection using single photons. Physica A: Stat Mech its Appl (2024) 649:129974. 10.1016/j.physa.2024.129974
29.
TamilselviPLathikaVJayachitraSArunkumarSBalasubramaniMKalaichelviV. Secure multi-party computation for collaborative data analysis in network security[C]. In: 2024 international conference on intelligent and innovative technologies in computing, electrical and electronics (IITCEE). IEEE (2024). p. 1–5.
30.
LiuJTianYZhouYXiaoYAnsariN. Privacy preserving distributed data mining based on secure multi-party computation. Computer Commun (2020) 153:208–16. 10.1016/j.comcom.2020.02.014
31.
SahinbasKCatakFO. Secure multi-party computation-based privacy-preserving data analysis in healthcare IoT systems[M]//Interpretable Cognitive Internet of Things for Healthcare. Cham: Springer International Publishing (2023). p. 57–72.
32.
NaiduPSKharatRTekadeRMendhePMagadeV. E-voting system using visual cryptography and secure multi-party computation[C]. In: 2016 international conference on computing communication control and automation (ICCUBEA). IEEE (2016). p. 1–4.
33.
PuHCuiZLiuT. An electronic voting scheme using secure multi-party computation based on secret sharing. Int J Netw Security (2021) 23(6):997–1004. 10.6633/IJNS.202111_23(6).06
34.
ZhangQXinCWuH. Privacy-preserving deep learning based on multiparty secure computation: a survey. IEEE Internet Things J (2021) 8(13):10412–29. 10.1109/jiot.2021.3058638
35.
TranATLuongTDKarnjanaJHuynhVN. An efficient approach for privacy preserving decentralized deep learning models based on secure multi-party computation. Neurocomputing (2021) 422:245–62. 10.1016/j.neucom.2020.10.014
36.
HuangYYuJWangDLuXDufauxFGuoHet alLearning-based fast splitting and directional mode decision for VVC intra prediction. IEEE Trans Broadcasting (2024) 2(1):681–92. 10.1109/tbc.2024.3360729
37.
CramerRDamgårdIB. Secure multiparty computation[M]. Cambridge University Press (2015).
38.
YaoAC. Protocols for secure computations. In: Proceedings of the 23rd IEEE symposium on foundations of computer science. Washington, DC, USA: FOCS’ 82 (1982). p. 160.
39.
HouMSunSYZhangW. Quantum private comparison for the socialist millionaire problem. Front Phys (2024) 12:1408446. 10.3389/fphy.2024.1408446
40.
HuangXZhangWFZhangSB. Efficient multiparty quantum private comparison protocol based on single photons and rotation encryption. Quan Inf Process (2023) 22(7):272. 10.1007/s11128-023-04027-9
41.
HouMWuYZhangS. Efficient quantum private comparison based on GHZ states. Entropy (2024) 26(5):413. 10.3390/e26050413
42.
HuangXZhangSXiaJ. Efficient quantum private comparison using locally indistinguishable orthogonal product states[C]. In: International conference on artificial intelligence and security. Cham: Springer International Publishing (2022). p. 260–73.
43.
HouMWuY. New quantum private comparison using bell states. Entropy (2024) 26(8):682. 10.3390/e26080682
44.
ShiRH. Quantum sealed-bid auction without a trusted third party. IEEE Trans Circuits Syst Regular Pap (2021) 68(10):4221–31. 10.1109/tcsi.2021.3103857
45.
QiuCZhangSChangYHuangXChenH. Electronic voting scheme based on a quantum ring signature. Int J Theor Phys (2021) 60:1550–5. 10.1007/s10773-021-04777-1
46.
WenJZhangZLanYCuiZCaiJZhangW. A survey on federated learning: challenges and applications. Int J Machine Learn Cybernetics (2023) 14(2):513–35. 10.1007/s13042-022-01647-y
47.
AhujaAKapoorS. A quantum algorithm for finding the maximum. arXiv preprint quant-ph/9911082. (1999).
48.
DurrCHoyerP. A quantum algorithm for finding the minimum. arXiv preprint quant-ph/9607014. (1996).
49.
HuangXZhangWZhangS. Practical quantum protocols for blind millionaires’ problem based on rotation encryption and swap test. Physica A: Stat Mech its Appl (2024) 637(10.1):129614. 10.1016/j.physa.2024.129614
50.
ChristensenRBPopovskiP. Private product computation using quantum entanglement. IEEE Trans Quan Eng (2024) 4:1–9. 10.1109/tqe.2023.3320052
51.
ChenYSituHHuangQZhangC. A novel quantum private set intersection scheme with a semi-honest third party. Quan Inf Process (2023) 22(12):429. 10.1007/s11128-023-04195-8
52.
HuangYChenJFengCChenR. Some families of asymmetric quantum MDS codes constructed from constacyclic codes. Int J Theor Phys (2018) 57:453–64. 10.1007/s10773-017-3578-1
53.
BealeSJWallmanJJGutiérrezMBrownKRLaflammeR. Quantum error correction decoheres noise. Phys Rev Lett (2018) 121(19):190501. 10.1103/physrevlett.121.190501
54.
HouMWuY. Two-party quantum private comparison protocol for direct secret comparison. Mathematics (2025) 13(2):326. 10.3390/math13020326
55.
LiJCheFWangZFuA. Efficient quantum private comparison without sharing a key. Entropy (2023) 25(11):1552. 10.3390/e25111552
56.
HouMWuY. Efficient quantum private comparison with unitary operations. Mathematics (2024) 12(22):3541. 10.3390/math12223541
57.
HuangXChangYChengWHouMZhangSB. Quantum private comparison of arbitrary single qubit states based on swap test. Chin Phys B (2022) 31(4):040303. 10.1088/1674-1056/ac4103
58.
GisinNFaselSKrausBZbindenHRibordyG. Trojan-horse attacks on quantum-key-distribution systems. Phys Rev A—Atomic, Mol Opt Phys (2006) 73(2):022320. 10.1103/physreva.73.022320
59.
DengFGLiXHZhouHYZhangZ. Improving the security of multiparty quantum secret sharing against Trojan horse attack. Phys Rev A—Atomic, Mol Opt Phys (2005) 72(4):044302. 10.1103/physreva.72.044302
60.
LiZHWangLXuJYangYAl-AmriMZubairyMS. Counterfactual trojan horse attack. Phys Rev A (2020) 101(2):022336. 10.1103/physreva.101.022336
Summary
Keywords
privacy-preserving, maximum value determination, local operations, fairness, security
Citation
Hou M and Wu Y (2025) Privacy-preserving maximum value determination scheme. Front. Phys. 13:1592890. doi: 10.3389/fphy.2025.1592890
Received
13 March 2025
Accepted
04 July 2025
Published
31 July 2025
Volume
13 - 2025
Edited by
Yuanyuan Huang, Chengdu University of Information Technology, China
Reviewed by
Bing Wang, Nanchang University, China
P. Muthulakshmi, SRM Institute of Science and Technology, India
Updates
Copyright
© 2025 Hou and Wu.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Min Hou, houmin@scujj.edu.cn
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.