qr code encoding

Existing privacy-preserving technologies struggle to achieve widespread adoption on mobile devices due to their complexity, particularly lacking zero-knowledge proof schemes that support offline verifiability. This work proposes zQR, a novel framework that integrates zkSNARKs with QR codes to enable offline verification on mobile platforms for the first time. To ensure auditability and non-repudiation, verification logs are immutably recorded on a blockchain, while large language models are leveraged to automatically generate verification circuits. Experimental results demonstrate that zQR efficiently supports both proof generation and verification within QR code version 19 at low error correction levels, achieving low latency and manageable on-chain gas costs. These findings confirm the practical feasibility of zQR as a deployable privacy-preserving solution for mobile environments.

This work proposes a lightweight authentication mechanism based on the Schnorr zero-knowledge protocol to address the challenge of secure and efficient identity verification in resource-constrained environments. The scheme integrates QR codes to transmit constant-size proofs (approximately 0.5 KB) containing a nonce and a timestamp, effectively mitigating replay attacks without revealing the private key. Leveraging 256-bit elliptic curve cryptography, both proof generation and verification are completed within milliseconds, substantially reducing computational and communication overhead. Notably, this study presents the first practical integration of QR codes with replay-resistant zero-knowledge proofs, enabling deployment on mobile and low-resource devices.

This work addresses the vulnerability of QR codes to forgery attacks—such as quishing—due to their lack of inherent security mechanisms, particularly in self-service terminal scenarios. The authors propose a dual-mode secure QR code architecture: a fully offline scheme leveraging EdDSA signatures, CBOR-encoded certificates, and ZLIB compression; and a hybrid Web PKI model integrating JWKS endpoints, a centralized trust registry, and URL fragments to support both offline verification and online dynamic validation. Notably, this is the first approach to embed Ed25519 signatures and CBOR certificates within standard QR code capacity limits, providing strong cryptographic integrity while enabling real-time key revocation, backward compatibility, and native camera scanning. The solution effectively mitigates forgery threats and resolves scalability challenges in large-scale IoT deployments.

Selecting appropriate zero-knowledge proof (ZKP) systems for privacy-preserving computation on resource-constrained edge devices remains challenging due to trade-offs in performance, security assumptions, and hardware compatibility. Method: This work presents the first end-to-end empirical evaluation of zk-SNARKs (Groth16) and zk-STARKs on a commodity ARM64 platform, measuring proof generation/verification latency, proof size, CPU bottlenecks, and trust model requirements. Contribution/Results: Groth16 achieves 68× faster proof generation and 123× smaller proofs than zk-STARK, but incurs slower verification and requires a trusted setup. Conversely, zk-STARK offers faster verification, post-quantum security, and no trusted setup—yet suffers from significantly higher generation cost and larger proofs. The study identifies critical impacts of low-level implementation choices and ARM64 microarchitectural features on real-world ZKP performance. It further proposes application-driven ZKP selection guidelines tailored for edge deployment, providing empirically grounded insights and practical recommendations for deploying privacy-enhancing technologies in constrained environments.

In cloud environments, server-computed results are vulnerable to tampering, while existing zkSNARKs incur high constraint counts and substantial proof-generation overhead for fundamental operations such as matrix multiplication. Method: This paper proposes zkVC, a zero-knowledge verifiable computation framework that introduces Constraint-Reduced Polynomial Circuits (CRPC) and Prefix-Sum Queries (PSQ) — the first of their kind — to drastically compress R1CS size and accelerate verification. zkVC integrates polynomial commitments with circuit-level optimizations to enable private, verifiable computation while preserving both algorithmic and data privacy. Contribution/Results: Experiments demonstrate over 12× speedup in zero-knowledge proof generation for matrix multiplication, enabling millisecond-scale proof generation on standard servers and sub-millisecond client-side verification. The framework is open-sourced.

Practical zero-knowledge proof (ZKP) systems have been largely confined to NP problems, leaving PSPACE-complete problems—such as Quantified Boolean Formula (QBF) evaluation—without efficient ZKP constructions. Method: This paper introduces the first practical ZKP protocol for QBF, built on a novel integration of quantified resolution (Q-Res) proofs and winning strategies, both encoded simultaneously via polynomial arithmetic encoding. The design combines lightweight arithmetic circuit verification with a customized zero-knowledge protocol to achieve low-overhead, end-to-end verifiability. Contribution/Results: Our approach breaks the NP barrier, delivering the first end-to-end practical ZKP for a PSPACE-complete problem. Evaluated on the QBFEVAL benchmark, 72% of instances complete Q-Res verification within 100 seconds, and 82% support winning-strategy verification. This work establishes a new foundation for zero-knowledge verification of high-complexity computational classes.

This work proposes a novel framework that introduces zk-SNARK zero-knowledge proofs into autonomous vehicle V2V/V2I communication to enable verifiable and trustworthy interactions while preserving privacy. Integrated within the perception stack, the framework efficiently generates and verifies integrity proofs of perception and decision-making computations without revealing sensitive sensor data, model parameters, or internal system states. Experimental results demonstrate that proofs can be generated in under 8 ms and verified in approximately 1 ms, achieving real-time performance suitable for vehicular applications. The design supports cross-platform interoperability and regulatory compliance, and the implementation has been publicly released as open-source software.

This work proposes the first watermarking mechanism for language models that simultaneously achieves unforgeability, recoverability, robustness, and undetectability. Existing approaches are vulnerable to false attribution attacks, wherein adversaries generate watermark-free text that is erroneously flagged as watermarked. To address this, the authors introduce a robust (recoverable) digital signature scheme based on attribute-preserving hash functions, enabling effective verification of texts that are semantically or syntactically close to the original while preventing forgery from distant inputs. The method guarantees that any detected watermark must originate from a specific model and can be traced back to its exact source generation, thereby significantly enhancing the reliability of content attribution and enabling fine-grained provenance tracking for generated text.

This work addresses the challenge of detecting quishing attacks—sophisticated QR code-based threats that evade conventional defenses—by proposing ALFA, a novel detection framework. ALFA reconstructs the QR code grid through binarization and integrates a pioneering FAST module for error correction alongside structural feature extraction, enabling a pre-trained model to assess the legitimacy of the code and prevent malicious redirection at its source. Moving beyond the limitations of vision-based deep learning approaches, ALFA effectively handles adversarial QR codes featuring color variations, geometric distortions, and other evasive manipulations. Experimental results demonstrate that ALFA achieves a remarkably low false negative rate of only 0.06% on synthetic datasets and significantly outperforms mainstream mobile scanning tools in real-world scenarios, offering both high accuracy and robust reliability.

This work systematically examines the structures of classical and quantum error-correcting codes in storage and communication, elucidating their deep connections to mathematical and physical objects such as sphere packings, lattices, combinatorial designs, group theory, and quantum phases of matter. The project pioneers a handbook-style integration of the error-correction knowledge base, employing a taxonomic framework that unifies information theory, algebraic coding theory, and interdisciplinary mapping techniques. Codes are organized structurally according to symbol types, enabling coherent classification and cross-referencing. The resulting resource not only serves as a rigorous and comprehensive reference but also empowers researchers to trace interrelationships among codes and inspire novel discoveries, thereby addressing a critical gap in systematic synthesis and cross-domain linkage within the field.