Card-based cryptography uses physical cards to realize cryptographic protocols, such as secure multiparty computation. At the last SecITC, card-based arithmetic operations using integer commitments were proposed with their application to statistical processing. In statistical processing, the statistical value increases as the number of data points increases, so the number of cards required for processing also increases. To address this issue, this paper introduces a new representation of commitments, named floating-point numerical commitments, for floating-point numbers. The floating-point numerical commitments allow us to represent a wide range of numbers, including negative and arbitrary-precision integers. We also propose a protocol that converts a conventional integer commitment to a floating-point numerical commitment and protocols for arithmetic operations using floating-point numerical commitments as inputs. We then discuss that our new representation and operations enhance the application of card-based cryptography to statistical processing.
Modern Windows kernel drivers, traditionally written in C, are prone to memory-safety bugs that could lead to security exploits and system instability. This paper evaluates the use of the Rust programming language for Windows driver development, by building safe abstractions around native kernel structures and APIs. A ProcMon-like driver has been developed to capture filesystem events and send them to user-mode components. We document the complete build and debug workflow, analyze crash diagnostics across Rust and C implementations, and evaluate runtime overhead through micro-benchmarks.
Matchmaking encryption (ME) is a generalization of attribute-based encryption in which both the sender and the receiver (each with their own attributes) can specify access policies for the other party. While prior work has clarified privacy and authenticity for ME, achieving adaptive privacy has typically required non-standard assumptions such as indistinguishability obfuscation (iO) or intricate selective-to-adaptive upgrades under strong assumptions, leading to impractical parameters. Moreover, the match condition is known to imply iO. In this work, we focus on the “mismatch” condition and give a simple construction of adaptively secure ME from witness encryption (WE), a weaker assumption than iO. The resulting scheme reduces reliance on strong assumptions while preserving the core functionality of ME, advancing the feasibility of practical ME under weaker assumptions than previously known schemes.
Quantum key distribution (QKD) offers a secure mechanism for sharing encryption keys, grounded in the principles of quantum physics, and remains resilient even in the presence of quantum computing capabilities. However, a fundamental challenge in QKD is authentication, which currently depends on classical cryptographic techniques. Post-quantum cryptography (PQC) and public key infrastructure (PKI) are frequently proposed for QKD authentication due to their scalability and efficient node verification, which eliminates the need for pre-shared key pairs across individual links. The practicality of PKI, however, hinges on the current assumption that no known attacks can significantly accelerate the compromise of underlying post-quantum digital signatures, and that public-private key pairs can be refreshed at a manageable rate. This paper investigates a scenario in which post-quantum digital signature algorithms offer only transient security, valid for a limited number of generated signatures. This work analyzes the volume of authentication key renewal data that would need to traverse the network under such conditions to preserve algorithmic security, demonstrating that digital signatures lose their advantage when their usage period is severely constrained. To address this limitation, a novel approach is proposed: integrating PKI with key recycling in QKD reduces the volume of authentication key renewal data required in QKD networks. The security analysis reveals that, given the stringent requirements of post-quantum digital signatures, the overall authentication security is likely to be dominated by the inherent vulnerabilities of QKD sessions, particularly when conventional security parameters on the order of ∼ 10^-10 are employed. These findings provide insights for the future-proof design of QKD networks.
The increasing reliance on Internet of Things (IoT) infrastructures has ampli-fied the need for robust intrusion detection and prevention systems (IDS/IPS). Artificial intelligence (AI), particularly deep learning, has been widely adopted to enhance detection accuracy and adaptiveness in these environments. Existing AI-based approaches demonstrate strong perfor-mance, yet challenges remain in terms of scalability, resilience to novel at-tacks, and the handling of highly complex feature spaces. Quantum Neural Networks (QNNs) represent a new computational paradigm with the poten-tial to address some of these limitations by exploiting properties such as superposition and entanglement to capture correlations beyond classical models. This paper examines the current state of AI-driven IDS/IPS for IoT networks and analyzes the potential impact of introducing QNN capabili-ties. The discussion considers architectural implications, possible perfor-mance gains, and the constraints of present-day noisy intermediate-scale quantum (NISQ) devices. Rather than proposing a concrete hybrid imple-mentation, the paper focuses on mapping how QNN concepts could com-plement or extend existing AI-based frameworks for IoT security. The anal-ysis highlights both the opportunities and the open challenges of integrat-ing quantum-inspired methods into intrusion detection, contributing to the broader discourse on next-generation approaches to securing IoT systems.
