Zero-knowledge proofs (ZKPs) enable a prover to convince a verifier of knowledge of a secret without revealing it. The ZKP for the square-root problem has many applications in network and cloud security, such as user authentication and privacy-preserving cloud storage auditing. Classical protocols for the quadratic residuosity (square-root) relation require multiple iterations to reach negligible soundness error, incurring latency and communication costs that are critical in cloud settings. This paper proposes a new single-round zero-knowledge proof (SR-ZKP) for the square-root problem that achieves the same soundness as iterative schemes by increasing the challenge length. The protocol requires only one execution of a 4-message protocol (request, commit, challenge, response) and can be transformed into a one-message non-interactive ZKP via the Fiat–Shamir heuristic. The completeness, soundness, and zero-knowledge properties of the proposed scheme are formally proven. The results of this study show that the proposed protocol can achieve approximately \(97\%\) reduction in communication overhead and latency, when compared to an 80-round iterative ZKPs with RSA modulus n of size 2048 bits. This provides a substantial advantage for cloud applications.
A Modular DSP Architecture for Extreme-Precision Computation of π Author: José Ignacio Peinador SalaContact: joseignacio.peinador@gmail.comORCID: 0009-0008-1822-3452 🎯 TL;DR: What's This About? Problem: Calculating π at extreme precision hits a "Memory Wall" — parallel algorithms choke on shared memory access. Breakthrough: We discovered that π's calculation can be decomposed using modular arithmetic (mod 6), creating 6 independent computation channels with zero inter-thread communication. Key Insight: This decomposition is grounded in a formal isomorphism with polyphase filter banks in Digital Signal Processing (DSP), a bridge between number theory and engineering established in our companion work. Result: ✅ 100 million digits of π computed with just 6.8 GB RAM (95% parallelisation efficiency) ✅ Shared-Nothing architecture with strictly isolated memory per channel ✅ Stride-6 transition leaf with exact phase correction, compressing recursion depth by 2.6× ✅ Open-source implementation in Python/gmpy2, executable on Google Colab's free tier Why it matters: This architecture transforms an intrinsically memory-bound problem into a CPU-bound one, enabling near-linear scaling on commodity hardware without specialised HPC infrastructure. 📖 Executive Summary This repository hosts the reference implementation and experimental validation of the Hybrid Stride-6 architecture for extreme-precision computation of π. The architecture exploits the arithmetic structure of the Chudnovsky series by decomposing it into six independent modular channels, each processed by a dedicated worker with its own memory space. The decomposition is not an ad hoc optimisation but rests on a rigorous mathematical foundation: the polyphase isomorphism between modular arithmetic on ℤ/6ℤ and multirate signal processing. This isomorphism guarantees perfect reconstruction (no information loss across channels) and orthogonality (no inter-channel interference). The architecture is validated through the 100M Barrier Run: computing 10⁸ digits of π on a resource-constrained Google Colab instance (2 vCPUs, 12 GB RAM) in under 20 minutes, with 95% parallelisation efficiency and a sustained throughput of over 83,000 digits per second. 🏆 Key Contributions 🔬 Theoretical Foundations (Summarised from Companion Work) Polyphase Isomorphism: Formal proof that modular decomposition of integer-indexed series is equivalent to polyphase decimation in DSP Hexagonal Lattice Connection: Geometric motivation via the A₂ lattice (densest circle packing in the plane) Perfect Reconstruction Guarantee: Mathematical proof that the six channels recombine without aliasing or leakage ⚡ Computational Architecture Shared-Nothing Design: Six independent Python processes with strictly isolated memory spaces Stride-6 Transition Leaf: Processes blocks of 6 consecutive terms in a single operation, reducing recursion tree depth by log₂6 ≈ 2.585 Critical Phase Correction: Direct accumulation of the linear term B(k) prevents off-by-one-stride phase errors 📊 Experimental Validation 100M Barrier Run: 100 million digits computed on 12 GB RAM with 95% parallel efficiency Orthogonality Verification: ℓ² norm of channel terms matches norm of original series to machine precision Reference Comparison: All 10⁸ digits match y-cruncher reference values exactly 📈 Performance Highlights 🚀 "The 100M Barrier Run" — Extreme Validation Metric Result Significance Digits Calculated 100,000,000 Exascale-capable architecture Total Time 1,194.32 s (19.90 min) Sustained performance on cloud hardware Parallel Efficiency 95% (1.90× speedup) Near-linear scaling on 2 cores Peak RAM Usage ~6.8 GB Runs within 12 GB Colab limit Throughput 83,729 digits/second Competitive with optimised implementations Numerical Integrity Bit-exact match with y-cruncher Zero cumulative error 🏗️ Architectural Comparison Aspect Monolithic Binary Splitting Hybrid Stride-6 (This Work) y-cruncher (State-of-Art) Memory Pattern Contiguous, saturates bus Local per core, optimises cache Sequential disk I/O Parallel Model Fine-grained synchronisation Embarrassingly parallel (6 processes) Optimised with locks Scalability Memory-bound CPU-bound, linear to 6 cores Disk-speed limited RAM Requirement Entire dataset in memory Working set reduced 6× Uses disk as RAM Design Philosophy Maximise single-thread speed Maximise resource efficiency Maximise absolute speed 🚀 Quick Start & Reproduction 1. Instant Online Experiment (Recommended) Click above to run the complete experimental validation in Google Colab — no installation required! 2. Key Experiments to Reproduce The companion notebook provides step-by-step reproduction of all manuscript claims: Theoretical Foundation: Verify the polyphase decomposition and energy conservation Stride-6 Algorithm: Test parallel computation with arbitrary precision (100k digits) 100M Barrier Run: Reproduce the full-scale benchmark (requires ~7 GB RAM) Performance Analysis: Measure speedup and parallel efficiency ⚙️ Technical Implementation Details The "Stride-6" Computational Engine Unlike conventional Binary Splitting (processes terms individually), our engine implements a compressed transition leaf that calculates the aggregate effect of 6 consecutive terms: def stride6_leaf(k_start): """Calculate compressed transition for block [k, k+5]""" P, Q, B_acc = 1, 1, 0 for m in range(6): n = k_start + m P_n, Q_n, B_n = compute_chudnovsky_term(n) P *= P_n Q *= Q_n B_acc += B_n # Critical phase accumulation T_leaf = Q * B_acc # Correct phase synthesis return P, Q, T_leaf Key Innovation: Direct accumulation of the linear term B(n) prevents phase drift, preserving arithmetic integrity at any scale. Shared-Nothing Architecture Each of the 6 workers operates in complete memory isolation: Independent address spaces (no shared memory locks) Local garbage collection (prevents heap fragmentation) Cache-optimised access patterns (maximises L1/L2 utilisation) Numerical Stability Guarantees Orthogonal decomposition — zero information loss (verified experimentally) Arbitrary precision backend (gmpy2) with proven numerical stability Exact phase correction in the Stride-6 leaf 📚 Citation & Academic Use If this work contributes to your research, please cite: @article{peinador2026modularDSP, title={A Modular DSP Architecture for Extreme-Precision Computation of π}, author={Peinador Sala, José Ignacio}, journal={Zenodo}, year={2026}, doi = {10.5281/zenodo.17768718}, url = {https://github.com/NachoPeinador/Arquitectura-de-Hibridacion-Algoritmica-en-Z-6Z} } The companion theoretical work establishing the polyphase isomorphism is: @article{peinador2026polyphase, title={Polyphase Isomorphism between Modular Arithmetic and Multirate Signal Processing}, author={Peinador Sala, José Ignacio}, year={2026}, publisher={Zenodo}, doi = {10.5281/zenodo.17680023} } 🌐 The Broader Research Programme This architecture is one component of a larger investigation into the computational and physical consequences of the ℤ/6ℤ modular symmetry. Related projects include: Polyphase Isomorphism: Formal mathematical proof of the isomorphism between modular arithmetic and DSP. Modular Substrate Theory: Unified framework for cosmology and hadronic physics. Topological State Preparation: Quantum register initialisation and dissipative protection via ℤ/6ℤ superselection. Common Thread: All projects leverage modular arithmetic (ℤ/6ℤ) as a fundamental organising principle across mathematics, physics, and computation. ⚖️ Licensing & Usage ✅ Academic & Research Use (Free) Available under PolyForm Noncommercial License 1.0.0: Permitted: Academic research, teaching, personal projects, non-commercial forks Requirements: Attribution, license preservation, non-commercial use ⛔ Commercial Use (License Required) Commercial applications require explicit permission, including: Integration into proprietary software products Commercial hardware benchmarking services SaaS platforms and cloud computing services 💼 For Commercial Licensing Inquiries:Contact: joseignacio.peinador@gmail.comSubject: "Commercial License Inquiry — Modular π Architecture" 🌟 Acknowledgments This independent research was enabled by: Infrastructure & Tools Google Colab for democratised computational resources Python ecosystem (gmpy2, NumPy, SciPy, Jupyter) for scientific computing GitHub for open collaboration infrastructure Data & References y-cruncher for validation benchmarks Digital Signal Processing community for foundational theory Community & Inspiration The open-source scientific community for collective knowledge advancement Independent researchers worldwide pushing boundaries outside traditional institutions Last updated: June 2026 | Version: 3.0 | Status: Actively Maintained
This study introduces a functional EEG-based Multi-Factor Authentication (EEG-MFA) system engineered for accessibility and security utilizing affordable consumer hardware. Our version uses the BioAmp EXG Pill ( |3,000) with Arduino UNO, which is far cheaper than standard biometric systems that need expensive medical-grade equipment (|50,000–|500,000). It gets 86.7% authentication accuracy when the signal is good.The system uses three authentication factors: a password (knowledge), a pattern (behavior), and an EEG biometric (inherence). This makes it more secure. We utilize One-Class SVM with RBF kernel (nu=0.1) for user modeling, which means we don’t have to collect fake data, which is a big problem when using biometrics. The system learns brain patterns unique to each user using just 3–5 enrollment recordings (12 seconds each) and a simple electrode setup (3 electrodes: forehead + ears).Recent improvements in open-source EEG gear have made it much cheaper. With devices like the BioAmp EXG Pill (around 3,000 rupees), OpenBCI boards (100–500 dollars), and NeuroSky MindWave (100 dollars), students can do projects and small-scale research that weren’t possible before with medical-grade equipment. This lower price makes it possible to look into EEG authentication outside of established labs, utilizing real-world consumer technology that has its own problems. Some of the most important new features are: (1) an adaptive learning mechanism that lowers the False Rejection Rate from 20% to 0% over five sessions while keeping the False Acceptances at zero; (2) a tolerance margin system (10%) that makes up for differences in electrode placement; and (3) a complete end-to-end implementation with FastAPI backend, PostgreSQL database, and Next.js frontend.When we tested with real consumer hardware, we found that the most important performance aspect was signal quality (electrode preparation). With the right setup, we got an 80% genuine acceptance rate and a 0% imposter acceptance rate. The 10% Equal Error Rate (EER) is higher than medical-grade systems (¡5%), but it shows that it is possible to use it for specialized security applications, educational research, and proof-of-concept deployments where cost is more important than accuracy.
The fast growth of Internet of Things (IoT) technologies has turned smart cities into big data ecosystems for intelligent mobility, energetic efficiency and public services. But this increasing reliance on IoT data raises significant privacy issues because of the perpetually gathered sensor readings, inter-organisational sharing and algorithmic analyses. In this paper, we focus on the state-of-the-art IoT data-sharing methods that preserve privacy and protect recent progress in preserving privacy while sharing data in the IoT personal record by preserving statistical value. It combines traditional approaches, including anonymisation, differential privacy, federated learning, secure multiparty computation and homomorphic encryption with new technologies (e.g., blockchain-enabled governance, edge intelligence or zero-knowledge proofs) (Nguyen et al., 2023; Alrawais et al., 2024; Lin & Kuo, 2025). The paper analyses the impact of hybrid architectures combining edge-cloud cooperation and decentralised access control for improving data protection, in terms of not losing performance or interoperability. Conclusions: Summary of the main findings, Barriers to Implementation. This paper identifies several ongoing barriers, including computational expense, related to past research. Personal data while preserving its analytical worth. It combines cutting-edge technologies like blockchain-enabled governance, edge intelligence, and zero-knowledge proofs with traditional strategies like anonymisation, differential privacy, federated learning, secure multiparty computation, and homomorphic encryption (Nguyen et al., 2023; Alrawais et al., 2024; Lin & Kuo, 2025). The study investigates how hybrid architectures that incorporate decentralised access control and edge-cloud collaboration can improve data security without compromising interoperability or performance. The results point to enduring obstacles, such as interoperability, computational overhead, and regulatory compliance, especially in urban settings with limited resources. In order to integrate privacy-by-design principles into IoT analytics for smart city governance, the study suggests a multi-layered conceptual framework. To maintain public confidence in urban digital transformation, this framework places a strong emphasis on open data policies, citizen consent procedures, and the incorporation of cutting-edge cryptographic techniques. The information adds to the current discussion on how to balance privacy and innovation in smart cities and provides guidance for system architects, legislators, and municipal IT leaders who want to adopt IoT responsibly.
Hassan Raza, Tsendayush Erdenetsogt, Muhammad Mohsin Kabeer, Muhammad Arsalan Aslam · 5 authors
The block chain technology has become a potential solution to improving security, privacy, and trust on contemporary data management systems. Conventional centralized systems are easily breached, tampered with and unauthorized access makes it necessary to have decentralized systems that cannot easily be tampered with. Block chain offers immutability, transparency, and cryptographic security and smart contracts offer automated access control and auditing. Sensitive information is safeguarded using privacy-saving methods, such as encryption, a zero-knowledge proof, and decentralized identity schemes. Scalability and collaboration are further increased with integration with cloud and big data systems. This review identifies the uses of Block chain, challenges and future research direction, which shows that Block chain is capable of changing the way secure and privacy-conscious data management is achieved.
Perkembangan teknologi komputer dan telekomunikasi meningkatkan efisiensi pengolahan data, namun menimbulkan tantangan keamanan, khususnya pada data rekam medis elektronik (RME) yang bersifat sensitif. Penelitian ini mengimplementasikan metode Zero-Knowledge Proof (ZKP) dan Revest Shamir Adleman (RSA) untuk meningkatkan keamanan dan privasi RME. ZKP memungkinkan pembuktian tanpa mengungkapkan informasi rahasia, sedangkan RSA menjaga kerahasiaan dan integritas data melalui enkripsi-dekripsi. Hasilnya, entropi data meningkat 24,53% (4,8314 menjadi 6,0165 bits/byte) setelah enkripsi RSA 2048-bit dengan padding OAEP berbasis SHA-256. Protokol ZKP metode Schnorr berhasil diimplementasikan tanpa membocorkan rahasia pengguna. Pengujian pada 100 pengguna simultan menunjukkan waktu respons rata-rata 1,8 detik dengan keberhasilan permintaan di atas 94%. Tantangan utama adalah beban komputasi autentikasi ZKP dan efisiensi saat jumlah pengguna bertambah. Integrasi RSA dan ZKP terbukti efektif meningkatkan keamanan, menjaga privasi, dan mempertahankan kinerja sistem RME.
Zero-knowledge proofs (ZKPs) have evolved from foundational interactive proof systems to highly efficient, scalable, and trusted-setup-free constructions powering today’s privacy-preserving and blockchain applications. The field began with the seminal works of Goldwasser, Micali, and Rackoff (GMR) and Goldreich, Micali, Wigderson (GMW) in the 1980s, which introduced interactive proofs, knowledge complexity, and showed that all NP languages admit zero-knowledge proofs. The 1990s brought non-interactive ZK (NIZK) via the CRS model (Blum–Feldman–Micali) and practical sigma-protocols like Schnorr proofs, establishing foundational tools still used today. From the 2000s through early 2010s, research integrated pairings, PCPs, and cryptographic soundness, culminating in pairing-based NIZKs and early succinct argument systems. The SNARK revolution accelerated with QAP-based zk-SNARKs (Gennaro–Gentry–Parno), practical implementations like Pinocchio and libsnark, and the highly efficient Groth16 proving system that became a blockchain standard. Since 2018, the field has shifted toward transparent, universal, and post-quantum-secure systems. Innovations include Bulletproofs (short proofs without trusted setup), zk-STARKs (scalable and PQ-secure), PLONK (universal/updatable setup), and Halo/Halo2 enabling recursive proofs without trusted setup. These advances underpin modern Zcash deployments, zk-rollups, and privacy-preserving scaling systems across Web3. Overall, the ZKP landscape has progressed from theoretical constructs to practical, scalable, and secure systems central to modern cryptography and decentralized computation.
The rapid expansion of Internet of Things (IoT) devices poses significant challenges for traditional centralized identity and access management (IdM) systems, which suffer from scalability limitations, single points of failure, and notable privacy risks. Although blockchain technology presents a promising decentralized solution, its direct adoption is often constrained by limited transaction throughput, high operational costs, and the computational constraints of IoT devices. To address these issues, this study proposes and rigorously evaluates HybID-AC, a novel hybrid architecture for decentralized identity and access management, specifically designed for large-scale, heterogeneous IoT ecosystems. HybID-AC employs a dual-layer design that separates global trust anchoring from local execution. A highly scalable, feeless Directed Acyclic Graph (DAG)-based distributed ledger functions as a public anchor layer, registering W3C-standard Decentralized Identifiers (DIDs) and access policy hashes. High-frequency access control operations are handled off-chain at the edge layer, leveraging the DIDComm v2 peer-to-peer protocol, Attribute-Based Access Control (ABAC) for fine-grained policy enforcement, and Zero-Knowledge Proofs (ZKP) to preserve attribute privacy. Analytical results demonstrate that the HybID-AC architecture significantly improves latency and cost-efficiency compared to fully on-chain approaches, maintaining stable performance even as network scale increases. Additionally, a novel probabilistic model is introduced to provide a quantitative measure of the integral security risk of ABAC policies under potential attribute compromise. Overall, the study concludes that this hybrid architecture effectively addresses the inherent trade-offs of blockchain in IoT systems, delivering a secure, scalable, and interoperable framework that empowers devices with self-sovereign identity while ensuring privacy and security by design.
Mary Jesselyn Co, Bruce Mitchell, Lisa Jordan Powell
This paper describes the development and implementation of an innovative hotel carbon reduction simulation aimed at developing sustainability competencies in business students. Using the Harvard Business School "Net Zero" simulation, we assess how interactive simulation-based learning opportunities improve students' self-assessment of the eight sustainability competencies with a mixed-methods approach. The research examines critical gaps in knowledge of how simulation-based methods can develop integrated competency sets needed to solve complex sustainability challenges. The simulation assigns around 950 first-year management students as hotel managers tasked with achieving 50% emission reductions over seven years while maintaining financial performance. Students select from 29 sustainability initiatives across Energy, Purchasing, and Management categories, working within realistic constraints of carbon budgets, site-specific emission factors, and dynamic market conditions. Our comprehensive analytical design combines pre-post competency surveys with cluster analysis of strategic approaches, and qualitative analysis of learning reflections. Anticipated outcomes are enhanced competency development across all eight dimensions, with gains in systems-thinking, futures-thinking, and implementation competencies. The research aims to provide empirical proof for developing specific sustainability competences as well as demonstrating scalable approaches of integrating sustainability education into core business curriculum.
Owen Dugan, Garcia, Roberto, Ronny Junkins, Jerry Liu · 8 authors
The success of large language models (LLMs) can be attributed in part to their ability to efficiently store factual knowledge as key-value mappings within their MLP parameters. Recent work has proposed explicit weight constructions to build such fact-storing MLPs, providing an improved understanding of LLM fact storage mechanisms. In this paper, we introduce an MLP construction framework that improves over previous constructions in three areas: it 1) works for all but a measure-zero set of feasible input-output pairs, 2) achieves asymptotically optimal parameter efficiency matching information-theoretic bounds for some embeddings, and 3) maintains usability within Transformers for factual recall. Through our improvements, we 1) discover a metric on value embeddings that characterizes facts-per-parameter scaling for both constructed and gradient-descent-trained MLPs, 2) identify a simple encoder-decoder mechanism that empirically matches gradient-descent MLP facts-per-parameter asymptotics across all the inputs and outputs we test, and 3) uncover a fundamental tradeoff between an MLP's fact-storage capacity and its usability within Transformers. Finally, we demonstrate a proof-of-concept application of fact-storing MLPs: modular fact editing on one-layer Transformers by \textit{replacing entire MLPs at once}.
K.Kalaiselvi, Mohammad Musa Al-Momani, E.Sivajothi, T. Vijetha · 6 authors
Embedded systems are at the heart of critical infrastructure facilities in areas such as energy, transportation, healthcare, defense; where integrity, traceability, and auditability of data generated by the system are of great importance. However, existing data provenance security solutions for embedded settings are highly unsatisfactory because centralized design is vulnerable in embedded environment, they do not scale well and have poor privacy mechanisms. This work presents a novel method of combining immutable data provenance and anonymous blockchain solutions to solve these issues and promote the trustworthiness of embedded systems. The framework takes advantage of privacy-preserving cryptographic methods such as ring signatures, stealth addresses, and zero-knowledge proofs to support tamper-evident decentralized storage of data events without losing source privacy. It's designed to run efficiently under the resource constraints of the embedded platforms, to be low point compatible with low-power devices, without sacrificing the responsiveness of the system or the authenticity of the data. The architecture is designed for real time monitoring and auditability on distributed embedded devices that are installed in critical infrastructure networks. A lightweight consensus algorithm designed for embedded environments allows secure synchronization and validation of data without the need for the heavy computation of a public blockchain. The framework was experimentally validated through prototype implementation and simulation in multiple use-case scenarios, showing its effectiveness against data forgery, unauthorized access and provenance tampering. Performance evaluation demonstrates that the model is scalable, low latency and high throughput under restrained resource environments. This work demonstrated that, by building immutable and anonymous data provenance into embedded systems, in addition to increasing transparency, trustworthiness, and robustness of operation, it is also possible to lay the foundation for a novel class of secure, decentralized infrastructure monitoring tools suitable for adversarial deployments. Results demonstrate a robustness for deployment into actual applications with high-assured data traceability supported with privacy protection.
In today's digital age, cloud storage and computing have become indispensable. Resource-constrained clients such as individuals and small organisations increasingly rely on powerful servers to store, manage and process their data. However, outsourcing data to external servers leads to significant privacy concerns, particularly when dealing with sensitive information such as medical records, financial transactions, or personal data. Fully homomorphic encryption (FHE) is a cryptographic technique that allows computation over encrypted data. In secure outsourcing with FHE, a client sends encrypted data to a server, which can perform requested computations without accessing the original data. The server returns the resulting ciphertexts, which the client can decrypt to obtain the final output. Despite its strong privacy guarantees, the practical adoption of FHE is limited by two main challenges: efficiency, which arises from the substantial performance overhead of FHE; and integrity, which stems from the lack of mechanisms to verify the correctness of the outsourced computation. In this thesis, we contribute to addressing these challenges in three aspects. First, we optimise oblivious algorithms for use in FHE, achieving improvements in key performance metrics and accelerating both bootstrapping and a range of applications. Second, we build efficient privacy-preserving information systems based on FHE. These include (i) two private machine learning protocols, the k-nearest neighbour algorithm and decision tree evaluation, (ii) SQUID, a secure system for storing and analysing genotype-phenotype data, and (iii) a protocol for securely delegating zero-knowledge proof generation. Third, we construct verifiable secure delegation of computation through FHE techniques. We provide the notion of blind proofs to provide integrity guarantees and demonstrate its practicality using blind zkSNARKs, a concrete instantiation of blind proofs.
P. Prakash, Faheema Kattakath Sanil, Jeffrey Tom Shaji, Saravanan Palani · 5 authors
The adoption of privacy-preserving techniques in healthcare is significant, especially while handling sensitive medical information. Traditional machine learning approaches raise significant concern regarding privacy, regulations, and data protection. Federated learning has emerged as an effective machine learning approach that enables a group of local models to collaboratively train the global model by sharing their updates instead of sharing the sensitive medical data. Nevertheless, a significant issue with federated learning is its vulnerability to various attacks, including model corruption and data tampering. The authors propose a methodology for developing a secure and privacy-safeguarded collaborative learning model by integrating zero knowledge proof (ZKP) with federated learning (FL). The proposed RS-ZKP methodology utilizes Pedersen commitments within ZKP to verify feature importance, ensuring that they fall within specified bounds without disclosing the actual values. The methodology is validated on two benchmark datasets using metrics accuracy, precision, recall, and F1 score.
Gukanraj S, Jeeva Rekha R, Dharshan M, Mohan Murthy M · 7 authors
Environmental pollution poses a significant threat to public health and ecosystems, demanding advanced methods for real-time monitoring and source identification. Traditional IoT monitoring systems often fail to capture complex spatiotemporal patterns and raise privacy concerns. This paper introduces a robust, privacy-preserving IoT-based environmental monitoring framework integrating Times Net for temporal feature extraction and Spatio-temporal Graph Neural Networks (STAGE) for spatial relationship modeling. The system incorporates Federated Learning with Differential Privacy, Zero-Knowledge Proofs (ZKP) for authentication, and Post-Quantum Cryptography (CRYSTALS-Cyber) for blockchain-secured model updates. Experimental evaluation using real-world IoT data demonstrates a 93.4% prediction accuracy, a 12% privacy gain, and a 35% reduction in communication cost compared to traditional methods. The architecture is scalable, modular, and designed to support real-time, privacy-sensitive environmental monitoring in smart city applications.
The proposed privacy-preserving identity management framework in this research is the combination of blockchain technology, zero-knowledge proofs, and adaptive sharding, which improves security, scalability and regulatory compliance and cybersecurity system. In contrast to current blockchain-based identity mechanisms, the presented mechanism involves encryption of identity attributes and distributes them on dynamic permissioned shards with the ability to selectively disclose identity information without leaking sensitive information. Comparison to Sovrin SSI, uPort and Civic show off impressive results, with a$33-63$percent increase in the transaction rate, 32 percent reduction in the duration of verification and 27 percent less time to generate the proofs. Compliance is achieved by reducing the storage overhead during which the immutability is preserved with chameleon hash-based redaction. There is a huge probability of leakage of privacy and efficiency of the consensus is 97 percent, so identity operations are sound and resistant to tampering. The architecture even makes it possible to have safe cross-domain authentication within heterogeneous environment. These results reassure that the framework will provide secure, efficient, and privacy-compliant identity management and thus will be very useful in enterprise, government, and IoT-based cybersecurity environments.
Efficient extraction and integration of pharmacokinetic (PK) data from scientific literature is critical for informed decision-making in drug development. Prior knowledge of PK parameters, particularly from similar compounds, supports first-in-human dosing, parameter estimation, and compound screening, ultimately helping to reduce attrition in clinical trials. While recent natural language processing (NLP) efforts have focused on extracting PK data from unstructured text, these approaches often overlook more comprehensive PK information and essential contextual metadata, which are usually reported in tables. Despite the prevalence and value of these tables, no previous work has systematically addressed the automated extraction of PK data from them. This thesis presents a novel NLP pipeline for identifying, extracting, and structuring PK data from scientific tables. The work addresses a key gap by targeting tables as a rich and underutilised source of PK information. The thesis is structured around four main components. First, a classification system combining supervised learning and prompt-based approaches is developed to retrieve PK-relevant tables from full-text biomedical articles. Second, heuristic and neural named entity recognition approaches are designed to extract PK parameters and associated metadata from table cells, including dose, species, study population, route of administration, units, and other contextual qualifiers. Third, an entity linking pipeline, including rule-based and zero-shot methods, is developed to normalise extracted data to a standardised PK ontology. Finally, the full pipeline is utilised to construct a large-scale PK database from PubMed Open Access articles. The database is evaluated through systematic sampling and manual quality assessment, and proof-of-concept analyses demonstrate how the extracted data can be used to characterise literature-wide reporting trends and explore comparative pharmacological questions. The results of this thesis demonstrate that automated PK table mining is both feasible and scalable, significantly accelerating the curation of high-quality datasets for pharmacometrics modelling. This work presents new open-source annotated corpora, domain-specific NLP methodologies, and practical tools for structuring PK literature, thereby opening the door to scalable, data-driven approaches in early drug development.
With the rapid development of quantum computing technology, traditional encryption methods face severe security threats in multi-party privacy intersection protocols in federated learning. In this paper, we propose a new protocol based on post-quantum cryptography. Firstly, lattice-based homomorphic encryption and zero-knowledge proof technology are used to achieve key generation and parameter initialization against quantum attacks. Secondly, ciphertext data encoding is carried out to support homomorphic operations. Next, a zero-knowledge proof is used to verify the correctness of the ciphertext intersection calculation. Finally, the protocol is embedded into the federated learning workflow, adaptively adjusting the parameters. Experimental results show that the protocol achieves the NIST (National Institute of Standards and Technology) security level 3, and the privacy leakage rate is less than 1.2%, the communication and computational costs are controllable, and the protocol does not bring great influence to the accuracy of the federated learning model. The experimental results verify that the protocol can provide a reliable protection for the privacy of federated learning data in the quantum era.
We present ZK IR, a novel 32-bit instruction set architecture (ISA) specifically designed for efficient zero-knowledge proof generation using STARK protocols. Unlike existing zkVMs that adapt general-purpose ISAs like RISC-V, ZK IR is designed from first principles to minimize proving overhead while maintaining compatibility with modern compiler toolchains. Our key contribution is a rigorous analysis demonstrating that a pure 32-bit register architecture with software-based multi-precision arithmetic outperforms designs with wider registers or specialized field arithmetic units. We achieve approximately 2× reduction in constraint count compared to naive approaches. ZK IR uses the Baby Bear field (31-bit prime) with Plonky3 for proving, and provides an LLVM-based compiler infrastructure enabling developers to write ZK applications in Rust, C, and C++.
Certificate authentication in online systems is required to ensure integrity and authenticity and prevent forgery. Traditional blockchain-based approaches work with double-chain architecture without any privacy-preservation capability or pack whole certificates into a single chain and incur substantial storage overhead. In this study, we introduce a light-weight dual-blockchain architecture with an external verification and audit side-chain and an inner chain for offline storing certificates. Zero-Knowledge Succinct Non-Interactive Arguments of Knowledge (Zero Knowledge Proofs) are employed to sign hashes without revealing sensitive information, and threshold signatures are employed to sign certificates. Compared to the traditional single-chain and double-chain architecture, the proposed system realizes up to 30% lower latency and 25% higher throughput based on experimental results on 1,000–100,000 certificate dataset. These results indicate the efficiency, scalability, and privacy-preserving feature of the proposed solution, which can be applied to large-scale applications for certificate management.
Jiaxi Liu, Lin Sun, Tianyu Kang, Di Wu · 7 authors
Federated Learning (FL) enables model training on distributed devices while preserving data privacy. However, malicious clients can submit fabricated model updates to fraudulently obtain training rewards, a behavior known as free-rider attacks. Existing detection-based solutions analyze anomalies in model updates but lack direct evidence of local training, making it fail to fully prevent free-riders. To address this limitation, we propose zkVFL, a verifiable FL framework leveraging Zero-Knowledge Proofs (ZKP) to ensure the integrity of local training while preserving privacy. To reduce the computational overhead of proof generation in ZKP, zkVFL introduces two novel techniques: (i) anomaly-aware client sampling to selectively perform ZKP verification and (ii) A recursive ZKP protocol (ReMPoT), incorporating a pruning-based layer selection technique, reduces proof generation costs. Experimental results demonstrate that zkVFL improves the accuracy and convergence of FL training under free-rider attacks while significantly reducing the computational and memory overhead of proof generation on resource-constrained devices.
Permissionless blockchains have evolved beyond cryptocurrency into foundations for Web3 applications, decentralized finance (DeFi), and digital asset ownership, yet this rapid expansion has intensified privacy vulnerabilities. This study provides a comprehensive review of recent trends, emerging privacy threats, and mitigation strategies in permissionless blockchain ecosystems. We examine six developments reshaping the landscape: meme coin proliferation on high-throughput networks, real-world asset tokenization linking on-chain activity to regulated identities, perpetual derivatives exposing trading strategies, institutional adoption concentrating holdings under regulatory oversight, prediction markets creating permanent records of beliefs, and blockchain–AI integration enabling both privacy-preserving analytics and advanced deanonymization. Through this work and forensic analysis of documented incidents, we analyze seven critical privacy threats grounded in verifiable 2024–2025 transaction data: dust attacks, private key management failures, transaction linking, remote procedure call exposure, maximal extractable value extraction, signature hijacking, and smart contract vulnerabilities. Blockchain exploits reached $2.36 billion in 2024 and $2.47 billion in the first half of 2025, with over 80% attributed to compromised private keys and signature vulnerabilities. We evaluate privacy-enhancing technologies, including zero-knowledge proofs, ring signatures, and stealth addresses, identifying the gap between academic proposals and production deployment. We further propose a Secure Development Lifecycle framework incorporating measurable security controls validated against incident data. This work bridges the disconnect between privacy research and industrial practice by synthesizing current trends, providing insights, documenting real-world threats with forensic evidence, and providing actionable insights for both researchers advancing privacy-preserving techniques and developers building secure blockchain applications.
This paper presents an empirical investigation of textual and semantic cues for fake news detection using FAKES-XL, a multi-domain, multi-language benchmark with leak-proof splits. Current reports often conflate gains with source/topic leakage and rarely assess probability calibration, limiting deployability across sources and languages. The present study trained text-only, semantic-only, and fused models on five bundles spanning English, Spanish, German, Hindi, and Italian, with temporal/source-grouped, topic-disjoint, cross-lingual zero-shot, and entity-disjoint evaluations. The methodology incorporated precommitted textual features (n-grams, stylometry, readability) and semantic signals (contextual embeddings, discourse, knowledge and retrieval-based evidence), applied post-hoc calibration, and quantified uncertainty via stratified bootstrap. Outcomes included Macro F1, Area Under the Receiver Operating Characteristic (AUROC), Area Under the Precision-Recall Curve (AUPRC), and Expected Calibration Error (ECE), with per-source and per-language scorecards and latency profiling under deployment constraints ($<=50 ~\text{ms}$on GPU;$<=120 ~\text{ms}$on CPU). While numeric results are not reported here, the analysis quantified the marginal value of each cue family, ablated discourse/knowledge/retrieval components, and produced calibrated thresholds tuned on validation and frozen on test. The contributions are a controlled comparison under strict leakage guards and a calibration-first evaluation that informs threshold selection. These findings support practical moderation workflows by offering reproducible scorecards and deployment-ready operating points.