The increasing use of federated learning (FL) in healthcare IoT demands rigorous verification to ensure the correctness of remote model training without compromising patient data privacy. However, existing approaches either assume full trust in clients or introduce high computational and communication costs when integrating cryptographic guarantees. In this work, we propose a lightweight, privacy-preserving federated learning framework that integrates zk-SNARK-based verifiable training over a ring topology. Our system ensures that each clientâs model update and aggregation step can be independently verified without revealing sensitive data or requiring a central auditor. We design an efficient proof composition strategy (CGro16) tailored for chained convolution operations and commitment schemes optimized for healthcare models. We also introduce a matrix polynomial-based masking mechanism (MatProofs) to support zero-knowledge commitments for convolutional neural networks (CNNs).Experimental results on standard benchmarks (MNIST, CIFAR-100) show up to 47% reduction in proof generation time and 39% lower memory overhead compared to baseline zk-SNARK schemes. The protocol is also benchmarked on edge devices (Jetson Nano, Raspberry Pi), confirming its suitability for remote and wearable healthcare scenarios.
Driven by globalization and digitization, the Mobile Industrial Supply Chain Internet of Things (IoT) has gradually developed, utilizing mobile devices and IoT technologies to enable real-time monitoring and efficient responses across various stages. However, with the growing demand for high-frequency data exchange, the Mobile Industrial Supply Chain IoT faces significant challenges in data security, authentication, and privacy protection. This paper proposes a security authentication scheme based on blockchain and group key management, leveraging the decentralized and tamper-resistant features of blockchain, the privacy-preserving authentication method of Zero-Knowledge Proofs (ZKP), and a hierarchical key management mechanism based on binary key trees. This approach aims to enhance the security and scalability of Mobile Industrial Supply Chain IoT. The experimental section simulates scenarios such as dynamic node addition and key updates, evaluating the performance in terms of encryption, decryption, and key management efficiency, thus demonstrating its superiority in multi-party collaborative environments.
Suhail Adel Alansary, Sarah M. Ayyad, Fatma M. Talaat, Mahmoud M. Saafan
Abstract The rise of artificial intelligence (AI) revolutionized both cybersecurity defenses and cybercriminals' methods to exploit vulnerabilities. Cybercriminals continue to exploit previously undiscovered vulnerabilities, known as zero-day attacks, posing severe threats to cybersecurity. These attacks are particularly challenging to detect, as they target unknown weaknesses in systems before security teams can respond or act. Traditional intrusion detection systems (IDS) rely heavily on pre-existing attack signatures, making them ineffective against zero-day threats. Machine learning (ML) algorithms have recently become a promising solution for enhancing IDS capabilities by identifying anomalies and predicting potential vulnerabilities in real time. This review paper explores how cutting-edge AI techniques, specifically ML, DL, and federated learning (FL), are harnessed to counter zero-day attacks. AI is used to defend against cyberattacks that exploit vulnerabilities unknown to existing security software. This research explores different AI methods used in cybersecurity, analyzes the data used to train these AI models, and evaluates how well various algorithms perform in actual cyberattacks. Moreover, key challenges in deploying ML for zero-day detection are highlighted, including handling imbalanced data, generalization across diverse types of attacks, and the trade-offs between accuracy and computational cost. The paper outlines future research directions to enhance AI-based zero-day attack defenses and strengthen proactive cybersecurity strategies.
Naser Abbas Hussein, Jihene Khoualdi, Ilhem Abdelhedi Abdelmoula, Hella Kaffel Ben Ayed
Internet of Things (IoT) has gripped domains with this ubiquitous connectivity, in-themoment data collection, and autonomous decision-making. But rising numbers of heterogeneous, extremely constrained IoT devices pose serious concerns regarding data privacy, security, and trust management, drawing great attention into these areas in the academic field and on all sides. Thus, blockchain technology came into the limelight for strengthening security and privacy in IoT systems in a decentralized manner, giving the system immutability, transparency, and distributed trust. This study proposes a Systematic Literature Review (SLR) of blockchain-based approaches that aim to enhance the IoT applications' privacy and security, focusing chiefly on healthcare, supply chains, and smart cities. The review uses a structured methodology to find, select, evaluate, and synthesize relevant peer-reviewed studies published between 2018 and 2025 taken from major scientific databases such as IEEE Xplore, ACM Digital Library, ScienceDirect, SpringerLink, and Scopus. Articles were also examined to narrow the scope of study and set the subject. The selected studies are analyzed and classified based on their security goals (e.g., confidentiality, integrity, authentication), privacy-preserving techniques (e.g., anonymization, differential privacy, zero-knowledge proofs), blockchain configurations (e.g., public, private, consortium), and consensus mechanisms. The findings reveal a growing body of research applying blockchain to a wide range of IoT domains, addressing diverse application domains such as healthcare, smart homes, industrial IoT, and agriculture, and demonstrating its potential to enhance data integrity, access control, and authentication. However, the integration of blockchain in IoT also faces challenges such as scalability, latency, and resource overhead, especially in real-time and constrained environments. This review offers a comprehensive synthesis of the state-of-the-art, identifies current limitations and research gaps, and proposes future research directions for building secure, efficient, privacy-aware, and scalable blockchain-enabled IoT systems.
This paper proposes and experimentally validates a holistic security framework for distributed systems, combining blockchain-based passport identity verification with AI-driven dynamic trust management. The framework addresses two critical challenges in decentralized environments: ensuring verifiable digital identities and maintaining scalable, adaptive trust evaluation. In the identity layer, electronic passports are used to generate zero-knowledge proofs, allowing users to demonstrate specific attributes without exposing sensitive personal information. This mechanism provides strong Sybil resistance and aligns with Self-Sovereign Identity principles. The trust layer incorporates machine learning models to continuously monitor node behavior and update trust scores in real time, enabling the system to respond to anomalies and malicious activities dynamically. To evaluate the practicality and effectiveness of the proposed framework, we developed a prototype system and conducted experimental validation in a simulated distributed environment. The results confirm that the integrated approach enhances authentication assurance, improves trust coordination, and supports sustainable scalability through efficient consensus and computation mechanisms. This work offers a promising direction for securing blockchain, IoT, and other decentralized systems. Future efforts will focus on field deployment, cross-domain interoperability, and regulatory compliance.
The widespread adoption of big data and AI technologies has accelerated the advancement of intelligent medical diagnostics. However, the sensitivity of medical data poses a dual challenge of privacy leakage and computational inefficiency in cross-institutional collaboration. Traditional federated learning (FL) schemes struggle to balance privacy protection, model accuracy, and communication costs, particularly for real-time processing of high-resolution medical images. To address this, we propose CZ-FLMed, a privacy-preserving FL framework integrating CKKS fully homomorphic encryption (FHE) and zkSNARKs zero-knowledge proofs. The framework employs a customized Convolutional Neural Network (CNN) for medical image training, the CKKS segmented encryption strategy for reducing communication overhead, and the lightweight Groth16 protocol for secure identity verification. This enables efficient encrypted model aggregation and authentication. The experiments results conducted in this paper on Chest X-Ray pneumonia dataset and MNIST handwritten digits demonstrate that CZ-FLMed achieves 83.05 % test accuracy in pneumonia classification. Compared to Paillier encryption, it reduces communication costs by 92.84 % and improves encryption efficiency by 404 times. Thus, the framework balances model accuracy, computational efficiency, and privacy preservation, offering a practical solution for multicenter medical collaboration.
Li Xingchen, Zhou Zhang, Burra Venkata Durga Kumar
One of the most important properties of cold chain is that ensure that temperature-sensitive products such as food, medicine, and chemicals maintain quality and safety during transportation and storage. For traditional cold chain systems, most operations such as transportation and inspection were relying on manual inspection and decentralized systems, which are inefficient, error-prone, and lack transparency. Today, some studies have combined blockchain technology with the Internet of Things (IoT) to store various necessary supply chain data on the blockchain, thereby achieving the role of monitoring and review, providing a basic solution to these challenges. But there still some problems, for example, how to attribute the responsibility in the transportation process to individuals to achieve a precise accountability system? For example, know who is responsible for this leg of the shipment? who is responsible for receiving this shipment? Since the temperature and humidity data of the fruit may be constantly changing, how can you effectively detect whether these changes are justified so that you can respond effectively and in a timely manner to irregularities? Regarding above mentioned issues, in this paper, we propose a two-tier framework that combines biometric-based Zero Knowledge Proof (ZKP) authentication and AE-based AI anomaly detection. The authentication subsystem uses biometric data and personal information to generate credentials, which are verified by the ZKP stored on the chain. Meanwhile, the IoT device collects multisource sensor data processed by feature engineering, and detects temperature, humidity, and route anomalies via VAE model.
Financial decisions in production systems must satisfy a layered set of obligations: risk tolerance, regulatory compliance, fairness constraints, privacy requirements, and operational service levels.Most machine learning models optimize predictive objectives but treat policy and compliance as external checks.This separation creates avoidable failure modes: decisions that are accurate yet non-compliant, long audit cycles, and limited customer recourse.This paper proposes Policy-Carrying Decision Models (PCDMs): decision systems that emit not only an outcome (approve/decline/route) and calibrated confidence, but also a machine-checkable proof that the decision adhered to an explicit policy expressed in a domain-specific language (FinPol).At inference time, the model (and its surrounding decision logic) produces a decision receipt containing the outcome, explanations scoped to permissible disclosure, and a verifiable policy proof.Optionally, a zero-knowledge variant allows third parties to verify compliance without access to sensitive features or thresholds.
Abdullah Ayub Khan, Asif Ali Laghari, Hamad Al-Mansour, Leila Jamel ¡ 8 authors
The multimedia environment has undergone significant growth, particularly in the area of multimedia data and its migration to cloud platforms, which has raised issues about security, confidentiality, data integrity, and privacy protection. While Blockchain Distributed Ledger Technology (BDLT) offers decentralized trust and transparency the advent of Quantum Computing threatens classical cryptographic primitives, which make multimedia data increasingly vulnerable. This paper proposes a novel and secure framework that collaborates BDLT with quantum-resilient, mainly known post-quantum cryptographic schemes to ensure long-term data integrity and privacy preservation in cloud-based infrastructures. Due to this, the proposed solution enables secure, efficient, and transparent that helps in public auditing of multimedia content without compromising stakeholder confidentiality. It leverages Zero-Knowledge Proofs (ZKPs), lattice-based cryptography, and smart contract automation, which model fortifies data authenticity verification against quantum attacks. Simulation results illustrate the effectiveness of the proposed framework that achieves a 98.21% accuracy in data integrity verification, a 96.84% reduction in quantum vulnerability, and an 87.85% efficiency gain in auditing speed compared to classical BDLT-enabled platforms. In addition, privacy leakage in multimedia systems is reduced by 92.47% proving the frameworkâs robustness. This solution underscores the potential of synergizing BDLT, quantum secure cryptography, and cloud computing to build a future-proof solution for privacy-protected multimedia data management and public auditing.
Open access
Cloud Data Security Solutions
Advanced Steganography and Watermarking Techniques
S N Prajwalasimha, Nilesh Shelke, Dilip Kumar Jang Bahadur Saini, Amit Pimpalkar ¡ 6 authors
Federated Learning (FL) is a decentralized collaborative AI training paradigm that maintains privacy of the data. FL is still susceptible to security attacks, malicious clients, and model integrity issues. To mitigate these issues, we introduce a Blockchain-Enabled Federated Learning (BFL) system that incorporates decentralized ledger technology to provide tamper-evident model aggregation, transparent client engagement, and verifiable updates. The suggested BFL framework uses smart contracts to enable automated trust management, zero-knowledge proofs (ZKPs) to facilitate privacy-enhanced authentication, and an incentive mechanism based on tokenized rewards to promote honest engagement. We also propose an adaptive consensus protocol that maximizes blockchain overhead while preserving high scalability for real-world applications like cybersecurity, healthcare, and Industrial IoT (IIoT). Experimental results on benchmark datasets show that BFL dramatically improves model robustness against data poisoning and adversarial attacks with a 15-25% improvement in attack resilience over state-of-the-art FL methods. Our work presents a complete blueprint for secure, privacy-preserving AI and establishes a foundation for the next generation of decentralized intelligence.
E-voting systems are gaining ground in international space, considering the versatility embedded in them in terms of making voting far more accessible and less expensive in logistics. Such systems, however, pose much higher security issues regarding privacy, especially while using biometric details, such as facial recognition, to verify the voter. The paper describes an e-voting system based on the blockchain architecture coupled with advanced privacy-preserving techniques like Zero-Knowledge Proofs (ZKPs) and end-to-end encryption. Voters have shied away from adopting this voting system as it is ensured by decentralization, transparency, and tamper-proofness, all at the cost of their identity and biometric data. In addition to this, this blockchain-based system also ensures that no one control is there for the voting process by any particular entity; hence, it guarantees impartiality. Because of ZKPs, it is ensured that voter verification can take place without exposing sensitive personal data in nature. Unlike the traditional e-voting-on-a-blockchain models, ours includes a dynamic key rotation mechanism for further anonymity of votes, with a multi-layer encryption to protect voter credentials even against potential quantum computing threats. Further, this blocking-based system makes sure that not a single entity is controlling the entire process of voting, thus guaranteeing impartiality. Voter verification takes place through the means of ZKP without any sensitive personal information being revealed. As a result, end-to-end encryption keeps intact the integrity of a vote from the moment it is cast until it is finally counted, further building confidence in the election process. It turns out to be a safe and scalable solution for future worldwide adoption of evoting.
Na Wang, Yaning Wang, Jianwei Liu, Junsong Fu ¡ 5 authors
In the development of urban transportation, Intelligent Connected Vehicle (ICV) technology has become a key force in promoting intelligence, sustainability, and efficiency. In particular, the rapid development of electric vehicles (EVs) has led to a growing demand for charging. However, this process risks exposing the private payment information of EV users, including the location of charging stations, the moment of charging, and users' identities. Our paper proposes a blockchain-based distributed privacy-preserving payment scheme named âSecretChargeâ, designed to protect users' private information for ICV payment service scenarios. To meet the demands of large-scale applications of ICV, we propose an efficient group signature scheme based on the ElGamal signature, achieving high efficiency in key generation, signing, and verification algorithms. To conceal the payment information of EV users, we sign the data using our group signature scheme and use it as input for a zero-knowledge proof, ensuring that the information can be verified without being disclosed. To eliminate the dependence on third-party billing entities in traditional centralized schemes, we process system payments on the blockchain, achieving decentralization. Finally, our scheme is tested on Ethereum. The experimental results demonstrate our scheme's efficiency and usability, achieving privacy protection for user payment information in ICV.
With the increased usage of cyber-physical systems (CPSs) in different critical domains, there is an emerging need for sound mechanisms for security and privacy. A deep learning-based biometric authentication system using feature extraction integrated with zero-knowledge proof and blockchain-based storage techniques is proposed for a secure authentication system providing better assurance in data privacy. The proposed architecture consists of three modular layers: a biometric processing layer responsible for extracting discriminative features using the FaceNet model, a cryptographic layer transforming these features into keys and generating ZKP-based proofs, and a blockchain layer for immutable authentication results storage. The system was tested on the Database of Faces, which resulted in 100% classification accuracy, reliably 194 integrated cryptographic functions through proof generation and verification times averaging 1.884 ms and 4.062 ms, respectively. Its compact proof size of 340 bytes speaks to the efficiency of the system. While these results clearly confirm the potential of such a system to be actually deployed in CPS applications, future work will address those challenges related to real-world conditions, such as diverse environments and additional biometric modalities. The proposed system enables a scalable and secure framework for CPS applications where privacy, transparency, and reliability are paramount.
V. Ananthakrishna, Bajrang Lal, Chandra Shekhar Yadav
The increasing demand for security and privacy-preserving collaboration among healthcare institutions presents significant challenges in data sharing, consent enforcement, and diagnostic automation, especially considering emerging quantum threats. This paper introduces PQ-FedCare, an innovative federated system architecture that incorporates post-quantum cryptography, zero-knowledge proofs, and smart contractâgoverned diagnostics to facilitate verifiable and privacy-compliant clinical collaboration. The proposed framework supports decentralized identity validation, encrypted consent delegation, and encrypted rule execution across blockchain-connected healthcare nodes. Using CRYSTALS-Kyber and SPHINCS+ for quantum-resistant security and zk-SNARKs for proof generation, PQ-FedCare ensures zero data exposure while enabling real-time, cross-institutional medical decision support. Evaluation on real-world clinical datasets (MIMIC-III, TCGA, and GEO GSE12102) demonstrates superior performance over recent baselines in diagnostic accuracy (94.5%), privacy leakage (0%), and proof verification time (92 ms). Additional stress tests confirm the systemâs robustness against missing data and scalability across federated nodes. The findings establish PQ-FedCare as a forward-compatible infrastructure for secure, accountable, and future-proof federated healthcare diagnostics. The proposed work is particularly suited for high-stakes clinical environments demanding transparency, regulatory compliance, and resistance to quantum-era attacks.
With the emergence of quantum computing, traditional cryptographic methods used in blockchain systems face increasing risk. One such area of concern is the Layer-2 protocols zk-Rollups designed to improve scalability and privacy in platforms like Ethereum, which are heavily dependent on classical zero-knowledge proofs such as zk-SNARKs and zk-STARKs. These systems may be compromised by quantum algorithms. To address this, we propose a quantum-secure zk-Rollup model using Quantum Zero-Knowledge Proofs (QZKPs), implemented with IBMâs Qiskit simulator. The protocol uses quantum features like superposition and random basis selection to verify transactions without leaking private data. Simulation results confirm key properties: valid proofs are reliably accepted, while invalid ones are rejected. This demonstrates both the feasibility and future relevance of integrating QZKPs into blockchain systems for post-quantum security.
Traditional digital identity models suffer from certain vulnerabilities in terms of identity reusability and privacy, as well as a single point of failure. The emergence of the blockchainbased self-sovereign identity model holds promise for addressing these issues in traditional digital identity models. However, existing schemes not only fail to cover privacy preservation throughout the entire lifecycle of credential issuance, verification, and revocation but also present security and efficiency concerns in key rotation. In this paper, we propose a novel blockchain-based self-sovereign identity system and redesign its credential scheme and key rotation mechanism. By leveraging the PS signature and zero-knowledge proof, our scheme preserves the privacy of holdersâ private attributes when issuing and verifying credentials. Additionally, with the cryptographic accumulator, our scheme does not reveal any issued or revoked credentials. Furthermore, we propose a secure and efficient key rotation mechanism based on pre-generated key chains, which enables secure and efficient key rotation without relying on a timelock. Finally, we provide a security analysis and performance evaluation, demonstrating the security and practicality of our scheme.
Democratic institutions increasingly rely on verifiable digital trust to enable fair participation and evidence-based decisions. Truvry is a decentralized protocol that converts behaviour-based evidence (usage patterns, transaction integrity, peer attestations) into portable cryptographic proofs that remain independent of any single platform or identifier, allowing individuals to transfer trust capital across domains while preserving privacy. The current prototype is zero-knowledgeâcompatible; in this version we use hashed proof anchoring and field-level redaction (no zk-SNARK module is deployed), with configurable smart-contract verifiers. By decoupling trust from identity, Truvry widens citizen inclusion, mitigates gatekeeping bias, and supplies auditable inputs for AI-mediated governance. In prototype tests (n=112), end-to-end proof issuance averaged 3.7 s (fastest local 1.4 s), verifier parse+check averaged 1.8 s, and the current minimum anonymization entropy is 8.9 bits; gas costs for optional on-chain anchoring remained below US$0.02. All results are based on simulated user streams; a production pilot is planned.
Finetuning Large Language Models (LLMs) is a highly effective way to improve their performance on the specific domains that need expertise knowledge. However, fine-tuning very large models is prohibitively expensive. A trending solution is to train a much smaller adapter, dubbed LoRA, serving as a âpluginâ to the model. However, in an untrusted distributed edge computing environment, when a user of an open-source base model wishes to utilize LoRA weights provided by external contributors, it is crucial to ensure that the LoRA weights are correctly matched with the intended base model and that the LoRA computation process is executed correctly. In this paper, we present ZK-EDGELORA, an efficient zero-knowledge (ZK) protocol that allows the LLM adapter (LoRA, the prover) to convince the base LLM model (the verifier) of its computing pro- cess, without revealing any information apart from the fact that the LoRA computing process is true. In particular, by leveraging VOLE-based âcommit-and-proveâ style ZK protocol, our solution enables efficient batch verification of matrix operations while preserving privacy. The proposed ZK-EDGELORA can safely and efficiently validate the correctness of each LoRA module within 0.1 to 2.8 seconds, depending on the weight size of the LoRA layer, when applied to real-world medical adapters from HuggingFace. The protocol establishes a scalable trust framework for distributed LLM deployments, bridging the gap between performance and security in modular AI ecosystems.
Dan Ivanov, Tristan Freiberg, Shahabi, Shirin, Jonathan Gold ¡ 5 authors
DSperse is a modular framework for distributed machine learning inference with strategic cryptographic verification. Operating within the emerging paradigm of distributed zero-knowledge machine learning, DSperse avoids the high cost and rigidity of full-model circuitization by enabling targeted verification of strategically chosen subcomputations. These verifiable segments, or "slices", may cover part or all of the inference pipeline, with global consistency enforced through audit, replication, or economic incentives. This architecture supports a pragmatic form of trust minimization, localizing zero-knowledge proofs to the components where they provide the greatest value. We evaluate DSperse using multiple proving systems and report empirical results on memory usage, runtime, and circuit behavior under sliced and unsliced configurations. By allowing proof boundaries to align flexibly with the model's logical structure, DSperse supports scalable, targeted verification strategies suited to diverse deployment needs.
Open access
2 source records
Adversarial Robustness in Machine Learning
Physical Unclonable Functions (PUFs) and Hardware Security
This work presents a cryptographic protocol for secure multi-party verification that achieves com putational privacy while maintaining exceptional computational efficiency. The proposed Position Based Commitment Protocol (PBCP) introduces a position-dependent nonce mechanism combined with cyclic verification architecture, enabling se cure computation over private inputs without re vealing individual parameters. Unlike existing commitment schemes that require complex cryp tographic assumptions, computationally expensive zero-knowledge proofs, or extensive public key in frastructure, Fundamental innovation lies in adapt ing physical laws of fluid dynamics to create nat ural mathematical relationships where each verifi cation equation contains multiple unknowns, mak ing parameter extraction computationally infeasible while preserving verification integrity. The proto col preliminary analysis suggests O(n) communica tion complexity with O(n2) verification complexity, providing substantial improvements over traditional Byzantine Agreement protocols that require O(n3) message exchanges. Comprehensive security analysis reveals robust resistance against statistical attacks with complexity O(R3) where R represents the pa rameter range, complete immunity to timing attacks through blind submission mechanisms, and resilience against collusion attacks involving up to n/2 â 1 ad versarial parties. The protocolâs unique cyclic neigh bor verification creates an interdependent validation network that prevents individual parameter extrac tion while maintaining system-wide integrity through mathematical interdependence rather than crypto graphic assumptions.
Blockchain, as a decentralized and tamperproof distributed ledger technology, has gained wide attention in finance, Internet of Things and other fields since it was proposed by Satoshi Nakamoto in 2008. Identity authentication is the basic guarantee for cyberspace security, but traditional centralized identity management suffers from single point of failure, privacy leakage and poor interoperability. Blockchain-based identity authentication utilizes distributed trust mechanism and cryptography technology, which is expected to realize secure sharing and autonomous control of identity data. In this paper, we systematically sort out the infrastructure (network layer, consensus mechanism, etc.) and types of blockchain technology, and elaborate the supportive role of the combination of blockchain and cryptography (hashing, digital signatures, zero-knowledge proofs, etc.) for identity authentication. It focuses on an overview of the research progress on the improvement of public key infrastructure (PKI), biometric combination scheme, and the integration of decentralized identity (DID) and verifiable credentials (VC) in the blockchain environment, and analyzes its application examples in the scenarios of Internet of Things (IoT), smart grids, finance, healthcare, and education. This paper summarizes the current challenges and limitations of blockchain identity authentication, such as performance scaling, privacy protection, standards interoperability, key management, etc., and the possible future research directions, including more efficient consensus algorithms, zeroknowledge proof applications, cross-chain identity mutual recognition mechanisms, and improvement of policies and regulations.