Medical big data holds significant value in promoting precision medicine, disease prediction, and public health management. However, issues such as sensitivity, decentralization, and privacy security limit its in-depth application. This study proposes a collaborative computing framework based on blockchain and Apache Spark, aiming to address the challenges of privacy protection, cross-institutional sharing, and efficient analysis of medical data. By designing an access control mechanism based on smart contracts and an anonymization scheme utilizing zero-knowledge proofs, combined with Spark's distributed memory computing advantages, a secure and trustworthy platform for medical data analysis is constructed. Experiments demonstrate that this framework improves data processing efficiency by 3.5 times compared to the traditional Hadoop architecture on the MIMIC-III dataset, while also meeting HIPAA privacy standards. This study provides theoretical support and practical pathways for the application of "blockchain + big data" technology in the medical field.
Traditional Supervisory Control and Data Acquisition (SCADA) are prone to cyber attacks which makes it difficult in keeping data secure. As the traditional architecture relies on centralized storage, which makes them vulnerable to unauthorized access and manipulation. To tackle these security concerns, the study demostrates the integration of blockchain technology and Bulletproof, a zero-knowledge range proof technique into SCADA systems to securely store aggregated sensor value proof in the ledger. The study uses Hyperledger fabric, creating a transparent and tamper-proof record for audits and providing a more secure and trustworthy system. This approach reduces storage and computational overhead while ensuring that data remains private and within valid ranges without exposing sensitive details. By combining blockchain with zero-knowledge range proofs, the proposed solution enhances both security and efficiency in SCADA systems. It ensures that sensor data remains protected from tampering while keeping blockchain resources optimized. This makes industrial systems more secure, reliable, and ready for the future of automation.
Blockchain technology, a decentralized and immutable ledger, has transformed identity and access management (IAM) by enhancing security, privacy, and trust in digital ecosystems. Ensuring safe authentication and data integrity is made possible by its integration with sophisticated cryptographic techniques like zero-knowledge proofs (ZKPs) and public- key infrastructure (PKI). Other methods include verifiable credentials (VCs) and decentralized identifiers (DIDs). This paper provides a comprehensive analysis of blockchain-based IAM systems, comparing leading blockchain platforms, including Ethereum, Hyperledger Indy, IOTA, and IoTeX, in identity management. The role of blockchain in mitigating identity-related threats, such as identity theft and unauthorized access, is explored through decentralization, immutability, and smart contract automation. Additionally, key security enhancements, including cryptographic mechanisms that strengthen decentralized identity solutions and privacy-preserving authentication, are examined. The potential of blockchain to establish a self-sovereign identity framework that fosters trust, scalability, and security in digital identity ecosystems is highlighted, paving the way for the next generation of identity management solutions.
With the growth of the Internet of Things (IoT), millions of users, devices, and applications compose a complex and heterogeneous network, which increases the complexity of digital identity management. Traditional centralized digital identity management systems (DIMS) confront single points of failure and privacy leakages. The emergence of blockchain technology presents an opportunity for DIMS to handle the single point of failure problem associated with centralized architectures. However, the transparency inherent in blockchain technology still exposes DIMS to privacy leakages. In this paper, we propose the privacy-protected IoT DIMS (PPID), a novel blockchain-based distributed identity system to protect the privacy of on-chain identity data. The PPID achieves the unlinkability of identity-credential-verification. Specifically, the PPID adopts the Zero Knowledge Proof (ZKP) algorithm and Shamir secret sharing (SSS) to safeguard privacy security, resist replay attacks, and ensure data integrity. Finally, we evaluate the performance of ZKP computation in PPID, as well as the transaction fees of smart contract on the Ethereum blockchain.
Open access
Blockchain Technology Applications and Security
Network Security and Intrusion Detection
Advanced Steganography and Watermarking Techniques
A future-proof food system should make optimal use of the available resources. Recovering proteins from by-products of plant-based food processing can contribute to the growing global protein demand. However, the extractability of plant proteins is limited by various factors, such as the entrapment in cellular structures or preceding processing steps. Additionally, recovered proteins should be functional to allow their use in food applications. Against this background, this dissertation aimed to evaluate both single and combined strategies for improving the extraction of functional proteins from potato trimmings and soybean okara, two case studies relevant to the (Belgian) food industry. Potato trimmings are a by-product from potato fries production, and okara is the insoluble residue obtained during the production of soy-based drinks or tofu. Proteins in potato trimmings and okara were shown to have different inherent extractabilities. In potato trimmings, approximately 50% of the proteins were readily extractable, while 25% were aggregated, and the remaining 25% were physically inaccessible, likely trapped within cellular structures. The extractability of okara proteins was low, with 50% restricted by protein aggregation and the other 50% by physical inaccessibility. Therefore, a comparison of identical extraction processes between these by-products can provide insights into how protein extraction is influenced by differences in biomass composition and inherent protein extractability. Ultrasound-assisted protein extraction disrupted potato and soybean cells with increasing ultrasonication time. This increased the protein yield up to 98% for potato trimmings and up to 90% for okara. Short ultrasound treatments (2000 J/g fresh weight, about 5 min) were especially effective in solubilizing aggregated okara proteins, increasing the protein yield from <10% to about 50%. Potato trimming proteins obtained by extended ultrasonication showed a slower adsorption at the air-water interface due to induced protein aggregation, which was correlated to a reduced foamability. In contrast, increasing ultrasound times decreased the size of extracted okara protein aggregates, which positively affected their foaming properties. In general, okara proteins were bad foaming agents, likely due to their aggregated nature. In contrast, isolated potato trimming proteins had very good foaming properties. The mechanism behind foam stabilization of potato proteins, and its major fractions patatin and protease inhibitors, was therefore investigated more in-depth. The foaming properties of proteins isolated from potato trimmings were better than those of a blend of commercial potato proteins containing similar relative amounts of patatin and protease inhibitors. This was the case despite the fact that more proteins were insoluble in the isolates from potato trimmings than the commercial isolates. Likely, impurities also contributed to these superior foaming properties, given the lower purity of potato trimming protein isolates compared to the commercial isolates. Patatin and protease inhibitors were shown to both contribute considerably to the formation and stabilization of foams. Interestingly, a beneficial effect on the foam stability was found when patatin and protease inhibitors of various colloidal sizes were present. Generally, the importance of acknowledging and studying the heterogeneity in potato protein fractions for understanding their foaming properties was highlighted. Incubation of the by-products with cell wall degrading enzymes (at pH 5.0) increased the protein yield after subsequent alkaline extraction (at pH 9.0) from 57 to 76% for cellulase pre-treated potato trimmings, and from 17 to 35% for pectinase pre-treated okara. A short ultrasound treatment during the enzyme-assisted process could be used to increase the efficiency, but only when applied during the alkaline extraction step. When applied during the enzymatic incubation at pH 5.0, protein recoveries were not increased or even decreased, despite that the ultrasound treatment could induce additional cell disruption. This was attributed to induced protein aggregation when ultrasound was applied at pH values close to the protein isoelectric point. The combined enzyme pre-treatment and ultrasound-assisted alkaline extraction increased the yield up to 87% for potato trimmings and up to 54% for okara. The pectinase pre-treatment of okara allowed the extraction of a less aggregated protein fraction while the additional ultrasound treatment extracted more aggregated proteins. The molecular weight distribution of extracted potato proteins was similar regardless of the used process. Mechanical particle size reduction by ball milling more efficiently disrupted cellular structures in freeze-dried potato trimmings than in freeze-dried okara. However, for both matrices, the protein recovery at pH 9.0 was not remarkably increased despite the observed particle size reduction. The reason for this differed between the by-products. For potato trimmings, the ball milling treatment induced protein aggregation which reduced their solubility, achieving a net zero effect of the treatment on the protein yield. For okara, the inherent aggregated nature of the proteins limited their extractability, regardless of the level of cellular disruption. For both matrices, applying a short ultrasound treatment at the start of the alkaline extraction was effective in extracting the aggregated proteins. This resulted in consecutive increases in protein recovery with decreasing particle size, with increases up to 76% for potato trimmings and up to 64% for okara. The combined ball milling pre-treatment and ultrasound-assisted isolation process reduced the purity and foamability of the isolated potato trimming proteins. Extensive co-isolation of starchy compounds was avoided by adding starch degrading enzymes during protein isolation, leading to an improvement of the protein foaming properties compared to proteins isolated from potato trimmings that were not ball milled. This was due to a more efficient air-water interfacial adsorption, presumably due to the formation of soluble aggregates containing partially unfolded potato proteins. In general, potato proteins had good foaming properties regardless of the used process, and alterations were better understood by investigating changes in their structural and air-water interfacial properties. Okara protein isolates always had low purity and contained heavily aggregated proteins regardless of the used process, limiting their ability to stabilize gas cells in foams. To summarize, this dissertation led to several key conclusions. A first conclusion is that cell disruption alone does not guarantee increased protein extractability. The impact of the cell disruption method itself should be considered, as well as the inherent solubility of the proteins. A second conclusion is that rational combinations of treatments are more effective in improving the extraction efficiency compared to single treatments. Third, identical processes can have different effects depending on the substrate, both in terms of how they impact protein recovery as well as how they impact protein functionality. Therefore, knowledge on the extraction-limiting factors and inherent protein properties is required to tailor extraction processes to specific substrates.
This research investigates privacy protection mechanisms and data security policy optimization for blockchain-based digital rights management platforms to balance transparency with robust privacy protection. A comprehensive experimental framework was developed, integrating advanced cryptographic techniques with intelligent policy management systems. A multi-layered validation methodology employed formal verification, black-box/white-box testing, and stress tests to validate performance across security, efficiency, and usability dimensions. The implemented solution provided 99.99% security assurance while achieving a 47% improvement in processing efficiency through zero-knowledge proofs and homomorphic encryption. Transaction processing reached 3,750 TPS (peaking at 4,200 TPS), with 99.8% regulatory compliance and 99.9% automated policy conflict resolution. The research demonstrates significant advancements in blockchain-based privacy protection through novel cryptographic implementation and automated policy management, establishing a robust framework for secure digital rights management. This solution offers substantial value for content delivery networks, digital asset management systems, financial institutions, and government services where the balance between transparency and privacy is critical, while reducing compliance management costs.
Electronic Health Records (EHRs) are now a necessary component of contemporary healthcare, but managing them presents a number of security, privacy, and interoperability issues. In order to solve these issues, this study introduces a unique framework for EHR management that combines four cutting-edge technologies: blockchain, Zero-Knowledge Proofs (ZKP), Ciphertext-Policy Attribute-Based Encryption (CP-ABE), and InterPlanetary File System (IPFS). Our solution makes use of the Ethereum blockchain for transparent and safe record-keeping, IPFS for efficient and decentralized data storage, CP-ABE for fine-grained access control, and ZKP for private authentication. We offer computational proofs for important components together with a thorough security analysis utilizing formal verification tools like ProVerif and Tamarin Prover. Comparing our framework to other alternatives, the findings show that it provides stronger security guarantees, better privacy protection, and increased scalability. Our approach also defends against a broader variety of possible threats, such as man-inthe-middle attack, repudiation attacks, and side-channel attacks. This work opens the door for more effective and patient-cantered healthcare information systems by advancing secure and privacy-preserving EHR management.
Hao-Tse Chung, Shao‐Hung Cheng, Yu‐Jia Chen, Li‐Chun Wang
Emerging Blockchain-empowered Federated Learning (BCFL) technology combines the decentralized security of blockchain with the privacy protection of federated learning. BCFL addresses the issue of single points of failure in centralized systems, making it an increasingly popular solution. However, current consensus mechanisms, such as Proof of Work (PoW), Proof of Stake (PoS), and Practical Byzantine Fault Tolerance (PBFT), lead to challenges such as high computational costs and limited scalability. This paper proposes a Batch Zero-Knowledge Proof-based practical Byzantine fault-tolerant (BZ-BFT) consensus mechanism for BCFL to enhance efficiency and reliability. By integrating Zero-Knowledge Proof (ZKP), our approach enables the verification of the primary node's proposal without revealing information from other network nodes, thereby ensuring the credibility of the aggregated results. To address the high computational overhead associated with ZKP, we present a batch quantization preprocessing technique called BatchZKP. Our proposed BZ-BFT reduces initialization, proof generation, and verification time by$97.81 \%, 70.0 \%$, and 47.64 %, respectively, significantly boosting BCFL system efficiency and reliability. Additionally, our approach reduces communication complexity from$O\left(n^{2}\right)$to$O(n)$and enhances Byzantine fault tolerance to${1/2}$.
The Internet of Vehicles facilitates seamless Vehicle-to-Everything (V2X) communication, offering a myriad of services ranging from traffic management to data exchange and route scheduling. However, the existence of malicious Autonomous Vehicles (AVs) poses significant security and privacy threats to data communications and vehicle users, respectively. Therefore, it is crucial to verify the identity and preserve the privacy of AVs before offering V2X services within each vehicular broadcast domain. To address the aforementioned issues, a novel privacy-preserving lightweight Fast Reed-Solomon Interactive Oracle Proof of Proximity using polynomial commitment-based authentication protocol is presented. The AVs are initially registered with a trusted authority in this protocol. After that, they are authenticated by roadside units in their respective broadcast domains using a zero-knowledge proof-based challenge-response mechanism. As per the performance analysis, the proposed protocol surpasses state-of-the-art authentication protocols and achieves notable improvements of 19.43% in registration computation time, 50.96% in registration latency, 89.75% in authentication computation time, 14.97% in authentication latency, 97.42% in handover computation time, and 95.84% in handover latency, compared to other protocols. A qualitative security analysis is also carried out to prove that the proposed protocol provides anonymity, privacy, user verifiability, and untraceability features.
Dosun-Fuwari is one of Nikoli’s pencil puzzles. It is known that the generalized Dosun-Fuwari puzzle is NP-complete. Due to the inherent difficulty of the puzzle, solvers may often question whether a solution exists. Such questions highlight the need for a method that can verify the existence of a solution without revealing it, thereby preserving the puzzle’s challenge. In this paper, we propose a physical zero-knowledge proof protocol for the Dosun-Fuwari puzzle, which can be executed using 4mn + 2n cards. Here, m × n is the size of the instance of the puzzle.
This paper is concerned with a natural variant of the contact process modeling the spread of knowledge on the integer lattice. Each site is characterized by its knowledge, measured by a real number ranging from 0 = ignorant to 1 = omniscient. Neighbors interact at rate $λ$, which results in both neighbors attempting to teach each other a fraction $μ$ of their knowledge, and individuals die at rate one, which results in a new individual with no knowledge. Starting with a single omniscient site, our objective is to study whether the total amount of knowledge on the lattice converges to zero (extinction) or remains bounded away from zero (survival). The process dies out when $λ\leq λ_c$ and/or $μ= 0$, where $λ_c$ denotes the critical value of the contact process. In contrast, we prove that, for all $λ> λ_c$, there is a unique phase transition in the direction of $μ$, and for all $μ> 0$, there is a unique phase transition in the direction of $λ$. Our proof of survival relies on block constructions showing more generally convergence of the knowledge to infinity, while our proof of extinction relies on martingale techniques showing more generally an exponential decay of the knowledge.
This paper presents an innovative access control model for decentralized power systems by integrating zero-knowledge succinct non-interactive arguments of knowledge (zk-SNARKs). The proposed model addresses the growing need for secure, privacy-preserving management of access rights within distributed energy infrastructures. By leveraging zk-SNARKs, the system ensures that authorization can be validated without disclosing sensitive information, thereby enhancing both security and operational transparency. The model is rigorously analyzed through a combination of theoretical proofs and simulation experiments, demonstrating its effectiveness in mitigating unauthorized access and reducing computational overhead. Our findings indicate that the zk-SNARK-based approach not only fortifies the security framework of decentralized power systems but also facilitates scalable and efficient access management. This work provides a critical step towards the realization of robust, privacy-aware energy networks in the era of smart grids and distributed generation.
With the continuous advancement of intelligent algorithms, people have put forward new demands for data privacy protection, and the traditional privacy protection technology can not meet the needs of reality. To address this issue, a model safety verification method is proposed by combining European distance and cosine similarity constraint. The results showed that in the MNIST dataset and FMNIST dataset, the impact of three different attacks on the proposed research scheme showed a low value below 2%. In the face of malicious customers with different proportions, the impact of the MNIST and the FMNIST dataset are stable at a low range, and the maximum did not exceed 5%. The findings denote that the raised method has excellent performance in the defense model poisoning attack, and can effectively improve the robustness and privacy security in the federated learning system.
Blockchain technology is rapidly evolving, with scalability remaining one of its most significant challenges. While various solutions have been proposed and continue to be developed, it is essential to consider the blockchain trilemma -- balancing scalability, security, and decentralization -- when designing new approaches. One promising solution is the zero-knowledge proof (ZKP)-based rollup, implemented on top of Ethereum. However, the performance of these systems is often limited by the efficiency of the ZKP mechanism. This paper explores the performance of ZKP-based rollups, focusing on a solution built using the Hardhat Ethereum development environment. Through detailed analysis, the paper identifies and examines key bottlenecks within the ZKP system, providing insight into potential areas for optimization to enhance scalability and overall system performance.
Abstract: Automated smart contracts represent a paradigm shift in decentralized governance by integrating artificial intelligence (AI) with blockchain technologies to enhance security, scalability, and adaptability. Traditional smart contracts, while enabling trustless and automated transactions, often lack the flexibility to adapt to dynamic regulatory frameworks, evolving economic conditions, and real-time security threats. AI-powered smart contracts leverage machine learning, reinforcement learning, and predictive analytics to optimize contract execution, detect fraudulent transactions, and enable self-adjusting governance mechanisms in Decentralized Autonomous Organizations (DAOs). Additionally, AI enhances blockchain consensus mechanisms, fraud detection, and risk assessment in Decentralized Finance (DeFi) applications. Privacy-preserving technologies such as zero-knowledge proofs (ZKPs) and quantum-resistant cryptography strengthen the security and confidentiality of AI-driven smart contracts. This research explores the convergence of AI and blockchain, examining how intelligent smart contracts can automate legal compliance, enforce dynamic contract logic, and optimize transaction fees while maintaining transparency and decentralization. By integrating AI-driven decision-making, automated dispute resolution, and scalable execution models, this study provides a comprehensive framework for secure, efficient, and intelligent decentralized governance. Keywords: AI-powered smart contracts, blockchain automation, decentralized governance, reinforcement learning, fraud detection, decentralized finance (DeFi), zero-knowledge proofs, quantum-resistant cryptography, DAO optimization, legal compliance automation.
Artificial Intelligence (AI) is profoundly transforming cryptography by significantly enhancing cryptanalysis techniques and informing innovative cryptographic design approaches. This survey reviews recent advancements in applying deep learning methods to side-channel and differential fault analyses, demonstrating substantial improvements over traditional methods in attack efficiency, accuracy, and resilience. Additionally, it highlights breakthroughs such as neural differential cryptanalysis, which expand classical cryptanalytic boundaries. In cryptographic design, Generative Adversarial Networks (GANs) have successfully automated the creation of high-quality cryptographic primitives, particularly S-boxes. Furthermore, AI shows promise in post-quantum cryptography (PQC) by uncovering potential vulnerabilities and optimizing cryptographic parameters. Despite these advancements, challenges persist regarding data dependency, model generalization, and interpretability. Future research directions emphasize enhancing AI model explainability, creating standardized benchmarks, and integrating AI with emerging technologies such as quantum computing and zero-knowledge proofs.
Open access
Cryptographic Implementations and Security
Chaos-based Image/Signal Encryption
Physical Unclonable Functions (PUFs) and Hardware Security
Modular arithmetic, particularly modular reduction, is widely used in cryptographic applications such as homomorphic encryption (HE) and zero-knowledge proofs (ZKP). High-bit-width operations are crucial for enhancing security; however, they are computationally intensive due to the large number of modular operations required. The lookup-table-based (LUT-based) approach, a ``space-for-time'' technique, reduces computational load by segmenting the input number into smaller bit groups, pre-computing modular reduction results for each segment, and storing these results in LUTs. While effective, this method incurs significant hardware overhead due to extensive LUT usage. In this paper, we introduce ALLMod, a novel approach that improves the area efficiency of LUT-based large-number modular reduction by employing hybrid workloads. Inspired by the iterative method, ALLMod splits the bit groups into two distinct workloads, achieving lower area costs without compromising throughput. We first develop a template to facilitate workload splitting and ensure balanced distribution. Then, we conduct design space exploration to evaluate the optimal timing for fusing workload results, enabling us to identify the most efficient design under specific constraints. Extensive evaluations show that ALLMod achieves up to $1.65\times$ and $3\times$ improvements in area efficiency over conventional LUT-based methods for bit-widths of $128$ and $8,192$, respectively.
Aiming at the severe challenges of Intelligent Connected Vehicle (ICV) in the field of data security, this paper designs a four-layer ICV data security framework, which covers the whole process from data collection, processing, storage and sharing to privacy protection. In the security framework, three technical methods of blockchain, Zero-Knowledge Proof (ZKP) and Post-Quantum Cryptography (PQC) are integrated. Blockchain is used to provide distributed trust mechanism and tamper-resistant data storage. ZKP is used to protect data privacy and realize data verification without leaking sensitive information. PQC is used to enhance data encryption and authentication mechanism to resist quantum computing attacks. At the same time, the data sharing security strategy is formulated, including initialization and key generation, data transmission preparation, zero-knowledge proof generation and verification, data reception and verification, blockchain recording and consensus, access control and data usage, etc., to ensure the security and privacy protection of data during the sharing process. The simulation experiment data of delay time, throughput and privacy protection intensity in the simulation environment show that the scheme not only enhances data privacy protection and data integrity verification but also improves the system's ability to resist future quantum computing threats and provides a solid guarantee for the data security of intelligent connected vehicles.
Popular technologies such as blockchain and zero-knowledge proof, which have already entered the enterprise space, heavily use cryptography as the core of their protocol stack. One of the most used systems in this regard is Elliptic Curve Cryptography, precisely the point multiplication operation, which provides the security assumption for all applications that use this system. As this operation is computationally intensive, one solution is to offload it to specialized accelerators to provide better throughput and increased efficiency. In this paper, we explore the use of Field Programmable Gate Arrays (FPGAs) and the High-Level Synthesis framework of AMD Vitis in designing an elliptic curve point arithmetic unit (point adder) for the secp256k1 curve. We show how task-level parallel programming and data streaming are used in designing a RISC processor-like architecture to provide pipeline parallelism and increase the throughput of the point adder unit. We also show how to efficiently use the proposed processor architecture by designing a point multiplication scheduler capable of scheduling multiple batches of elliptic curve points to utilize the point adder unit efficiently. Finally, we evaluate our design on an AMD-Xilinx Alveo-family FPGA and show that our point arithmetic processor has better throughput and frequency than related work.
M Anjali, Sk. Ashraf Vali, J. Sai Divya, B. Lokesh
Online exams are a general aspect of education nowadays. However, they are still suffering with several security problems such as data breaches, result-changing cheating, and impersonation scams. Some of the modern systems utilize blockchain technology to improve security but come with problems such as being difficult to integrate, taking time to process, being costly, and lacking adequate privacy protection. This work introduces a blockchain-based structure that addresses these concerns. It utilizes smart contracts in a more improvised manner, employs zero-knowledge proofs to secure privacy. The system provides safe sharing of questions, decentralized storage of answers, and tamper-evident grading. It also has seamless integration with learning management systems (LMS). In comparison to traditional and blockchain-based examination systems, this method provides upgrades in security, efficiency, and speed. With the integration of decentralized technology and state-of-the-art security methods, this system provides internet-based exams that are secure, transparent, and economical.
Federated Learning (FL) has emerged as a promising paradigm in distributed machine learning, enabling collaborative model training while preserving data privacy. However, despite its many advantages, FL still contends with significant challenges -- most notably regarding security and trust. Zero-Knowledge Proofs (ZKPs) offer a potential solution by establishing trust and enhancing system integrity throughout the FL process. Although several studies have explored ZKP-based FL (ZK-FL), a systematic framework and comprehensive analysis are still lacking. This article makes two key contributions. First, we propose a structured ZK-FL framework that categorizes and analyzes the technical roles of ZKPs across various FL stages and tasks. Second, we introduce a novel algorithm, Verifiable Client Selection FL (Veri-CS-FL), which employs ZKPs to refine the client selection process. In Veri-CS-FL, participating clients generate verifiable proofs for the performance metrics of their local models and submit these concise proofs to the server for efficient verification. The server then selects clients with high-quality local models for uploading, subsequently aggregating the contributions from these selected clients. By integrating ZKPs, Veri-CS-FL not only ensures the accuracy of performance metrics but also fortifies trust among participants while enhancing the overall efficiency and security of FL systems.
Anne Broadbent, Alex B. Grilo, Nagisa Hara, Arthur Mehta
In a proof of knowledge (PoK), a verifier becomes convinced that a prover possesses privileged information. In combination with zero-knowledge proof systems, PoKs play an important role in security protocols such as in digital signatures and authentication schemes, as they enable a prover to demonstrate possession of certain information (such as a private key or a credential), without revealing it. A PoK is formally defined via the existence of an extractor, which is capable of reconstructing the key information that makes a verifier accept, given oracle access to any accepting prover. We extend this concept to the setting of a single classical verifier and multiple quantum provers and present the first statistical zero-knowledge (ZK) PoK proof system for problems in QMA. To achieve this, we establish the PoK property for the ZK protocol of Broadbent, Mehta, and Zhao (TQC 2024), which applies to the local Hamiltonian problem. More specifically, we construct an extractor which, given oracle access to a provers' strategy that leads to high acceptance probability, is able to reconstruct the ground state of a local Hamiltonian. Our result can be seen as a new form of self-testing, where, in addition to certifying a pre-shared entangled state, the verifier also certifies that a prover has access to a quantum system, in particular, a ground state; this indicates a new level of verification for a proof of quantumness.
Federated learning (FL) enables multiple participants to collaboratively train machine learning models while ensuring their data remains private and secure. Blockchain technology further enhances FL by providing stronger security, a transparent audit trail, and protection against data tampering and model manipulation. Most blockchain-secured FL systems rely on conventional consensus mechanisms: Proof-of-Work (PoW) is computationally expensive, while Proof-of-Stake (PoS) improves energy efficiency but risks centralization as it inherently favors participants with larger stakes. Recently, learning-based consensus has emerged as an alternative by replacing cryptographic tasks with model training to save energy. However, this approach introduces potential privacy vulnerabilities, as the training process may inadvertently expose sensitive information through gradient sharing and model updates. To address these challenges, we propose a novel Zero-Knowledge Proof of Training (ZKPoT) consensus mechanism. This method leverages the zero-knowledge succinct non-interactive argument of knowledge proof (zk-SNARK) protocol to validate participants' contributions based on their model performance, effectively eliminating the inefficiencies of traditional consensus methods and mitigating the privacy risks posed by learning-based consensus. We analyze our system's security, demonstrating its capacity to prevent the disclosure of sensitive information about local models or training data to untrusted parties during the entire FL process. Extensive experiments demonstrate that our system is robust against privacy and Byzantine attacks while maintaining accuracy and utility without trade-offs, scalable across various blockchain settings, and efficient in both computation and communication.