.png)
As quantum computing advances, its threat totraditional cryptographic protocols, especially for long-term encrypted data,becomes critical. This paper presents an agile cryptosystem designed to easethe transition from pre-quantum to post-quantum security by supportingefficient integration of post-quantum Key Encapsulation Mechanisms (KEMs). Ourapproach combines a CCA-secure KEM with robust Authenticated Encryption (AE),allowing only the encapsulated key to be updated during migration, withoutre-encrypting large data payloads—saving both computation and bandwidth. Weformalize cryptographic agility via an agile-CCA security model, ensuring thatneither the original nor updated ciphertexts leak information. A game-basedproof shows that the construction remains agile-CCA secure if the underlyingKEM and AE are individually CCA-secure in the random oracle model. The resultis a future-proof scheme that enables enterprises and cloud providers tosafeguard vast data volumes against emerging quantum threats with minimaldisruption.
Status: published
Full article here!
Quantum Key Distribution (QKD) networks arereceiving growing attention by researchers and funding agencies, with thelargest projects spanning hundreds of kilometers and large, multi-nationalinitiatives well under way. The growing scale of QKD networks poses some extrachallenges, due to the current impossibility of establishing end-to-endsecurity between nodes: for large distances, reliance on intermediate trustednodes is required, an aspect which increases the attack surface of QKDnetworks. A possible mitigation strategy for this problem lies in leveragingthe complexity of the network: by making use of multiple paths and applyingtechniques issued from cryptography and/or error-correcting theory, it ispossible to mitigate some attacks and force an adversary to compromise morenodes in the network. This paper discusses relevant approaches to leveragemulti-path aspects of QKD networks for increased security, analyzing thedifferent security assumptions, guarantees, and performance of differenttechniques, and presenting their applicability to the different networks.
Status: published
Full article here!
Attribute-based signatures (ABS) providefine-grained control over who can generate digital signatures and have manyreal-world applications. This paper presents a pair of fast ABS schemes: onefor Key-Policy ABS (KP-ABS) and another for Signature-Policy ABS (SP-ABS). Bothschemes support expressive policies using Monotone Span Programs (MSP), andoffer practical features such as large universe, arbitrary attributes, andadaptive security. Most notably, we provide the first implementation ofMSP-based ABS schemes and demonstrate that our schemes achieve the best-knownasymptotic and concrete performance in this domain. Asymptotically, keygeneration, signing and verification time scale linearly with the number of attributes;verification requires only two pairing operations. In concrete terms, for 100attributes, our KP-ABS scheme performs key generation, signing, andverification in 0.16s, 0.10s, and 0.13s, respectively; our SP-ABS schemeachieves times of 0.082s, 0.26s, and 0.21s for the same operations.
Status: accepted not yet published
Full article here!
Hybrid Homomorphic Encryption (HHE) is emerging as a practical alternative to fullyhomomorphic encryption by offloading computational overhead to the cloud.Despite the growing number of HHE schemes and implementations, no unifiedevaluation methodology currently exists. In this work, we present acomprehensive and reproducible benchmarking framework covering both standardand HE-friendly HHE schemes. We analyze and evaluate 19 open-source HHEframeworks, comprising 218 distinct ciphers' benchmarks, across diverse HElibraries and programming languages. To ensure fair comparison, we interpretthe main results under an HHE-128 security target, separating standardized orauthor-claimed 128-bit settings from below-target measurements. Ourcontributions include a unified repository, a language-agnostic benchmarkingtool, and detailed metrics on runtime and memory usage. The results offeractionable insights into the security-aware performance trade-offs of eachdesign and lay the groundwork for standardizing future HHE evaluations.
Status: accepted at 31stEuropean Symposium on Research in Computer Security (ESORICS) 2026, September14-18, 2026.
Full article here!
The BLS digitalsignature scheme, in particular its instantiation with the BLS12-381 curve, hasbecome a cornerstone of modern blockchain protocols such as EthereumProof-of-Stake, due to its unique and attractive characteristics (e.g., supportfor non-interactive signature aggregation). Recently, Cheng et al. (CHES 2025)demonstrated that the enormous Single-Instruction-Multiple-Data (SIMD)computing power of the Intel AVX-512 extensions, when combined with carefullydesigned vectorization strategies, can be effectively leveraged to speed up thecomputation of the optimal ate pairing on BLS12-381, a major component of BLS.This naturally raises the question of whether such SIMD-parallel processing canbe exploited more extensively to benefit the entire BLS signature scheme. Thepresent paper answers this question positively by presenting a highlySIMD-optimized BLS implementation using Intel AVX-512, especially theAVX-512IFMA instructions. In order to harness AVX-512 more efficiently for theperformance-critical operations of BLS, we explored a wide range ofoptimization options, including various formulas and vectorizationgranularities for elliptic curve arithmetic operations, scalar multiplication,and hash-to-curve, as well as the fine-tuning and flexible use of differentimplementations of the finite-field arithmetic. Benchmarking results collectedon an Intel Core i3-1005G1 ("Ice Lake") CPU show that our vectorizedBLS software using AVX-512 is at least 1.57 times faster than an x64 assemblyimplementation of the widely-used blst library.
Status: accepted not yet published.
Full article here!
The transition towards Industry 5.0 requiresarchitectures and workflows that extend beyond interoperability and automationto incorporate human-centricity, trustworthy Artificial Intelligence (AI),resilience, data sovereignty, semantic interoperability, Digital Twins, andcloud-edge continuum integration. While established models, such as theIndustrial Internet Reference Architecture (IIRA) and the Reference ArchitecturalModel Industrie 4.0 (RAMI 4.0), provide solid foundations for industrialdigitalization, they do not fully address emerging technological and regulatoryrequirements, including federated AI, agentic collaboration, Zero TrustArchitecture (ZTA), Data Spaces, and compliance with evolving Europeanregulations. This paper extends and adapts the IIRA to the requirements ofIndustry 5.0 through nine Architectural Capability Principles (ACPs)synthesized from the literature and emerging industrial initiatives. Theproposed framework systematically maps these principles to the IIRA viewpointsand provides a capability-oriented methodology for assessing the architecturalreadiness of Industry 5.0 ecosystems. The approach is validated through acomparative capability mapping of eleven European projects, identifyingrecurring architectural patterns, convergence areas, capability gaps, andpriorities for future standardization. The results show strong maturity incybersecurity, trustworthy AI, regulatory compliance, and cloud-edgetechnologies, while highlighting the need for greater integration ofhuman-centricity, semantic interoperability, and agentic collaboration. Bycombining an extensible architectural framework with evidence from large-scaleEuropean initiatives, this work provides practical guidance for evolvingindustrial reference architectures towards the standardization andimplementation of Industry 5.0 ecosystems.
Status: accepted at CloudCon 2026 through CSRN cluster.
.png){kind=icon}
.png){kind=icon}
.png){kind=icon}
The project co-funded under Grant Agreement No. 101190366 is supported by the European Cybersecurity Competence CentreFunded by the European Union.
Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or European Cybersecurity Competence Centre.
Neither the European Union nor the granting authority can be held responsible for them.
tagliato.png)