Institution profile

University of Padova

Academic institutioneurope · it
Official website
Research library523linked papers
Opportunities0open roles
Selected work

Representative Papers

5G NR Non-Terrestrial Networks: From Early Results to the Road Ahead

Jan 08, 2026arXiv.org

This study addresses the critical applications of 5G non-terrestrial networks (NTN) in remote coverage, emergency communications, and terrestrial traffic offloading. It systematically reviews the 3GPP NR-NTN standardization evolution from Release 18 to Release 20 and proposes an innovative architecture tailored for Release 20. Leveraging the ns-3 simulation platform, the work develops a high-fidelity satellite communication model—encompassing both low Earth orbit (LEO) and geostationary orbit (GEO) constellations—calibrated with 3GPP-standardized parameters and real-world scenario data. Comprehensive performance evaluations across diverse network configurations demonstrate that the proposed architecture achieves significant improvements in wide-area coverage and network resilience, thereby providing key technical enablers for integrated space-air-ground 5G networks.

2 citationsRead paper

A taxonomy of categories for relations

Feb 14, 2025arXiv.org

This paper addresses the lack of a unified classification framework for the structural properties of relational abstractions. Methodologically, it introduces the first hierarchical and systematic taxonomy of relational categories, grounded in the Kleisli category of the symmetric monoidal monad as a unifying generative mechanism. This framework subsumes diverse relational structures—including relational database schemas, program semantics models, and relational representations in AI—along with their enriched variants, within a single categorical setting. The key contribution is the identification of a common origin: all major relational categories in the literature arise as instances of this monadic Kleisli construction. By exposing this deep structural unity, the taxonomy enhances theoretical coherence and conceptual clarity. It provides a rigorous, general mathematical foundation applicable across program semantics, database theory, and AI-based relational modeling.

2 citationsRead paper

Adaptive partition Factor Analysis

Oct 24, 2024

Traditional factor analysis struggles to distinguish latent factors shared across multiple studies from study- or subgroup-specific sources of variation. To address this, we propose an adaptive multi-study joint factor model that employs a novel hierarchical shrinkage prior to induce sparsity and structural adaptivity in factor loadings. This is the first Bayesian framework to rigorously ensure identifiability of multi-study factor loadings while enabling unbiased estimation of subgroup-specific factors. The method flexibly infers hierarchical factor structures—from globally shared factors, to cross-study subgroups, down to fine-grained factors nested within individual studies. Simulation studies demonstrate estimation accuracy comparable to state-of-the-art methods, with substantially improved interpretability. Applied to avian co-occurrence and ovarian cancer gene expression datasets, the model successfully identifies robust cross-cohort biological signals and subgroup-specific driver factors.

2 citationsRead paper

A novel step-by-step procedure for the kinematic calibration of robots using a single draw-wire encoder

Feb 23, 2024The International Journal of Advanced Manufacturing Technology

In high-precision manufacturing, robotic positioning accuracy lags significantly behind repeatability—yet conventional calibration methods rely on expensive external equipment (e.g., laser trackers) or multiple sensors. Method: This paper proposes a novel, stepwise kinematic calibration approach that uses only a single draw-wire encoder for one-dimensional distance measurement. It establishes an end-to-end framework comprising pose-error decoupling modeling, Jacobian-based sensitivity analysis, and kinematic-constraint-driven iterative optimization—integrating geometric modeling, least-squares parameter identification, and constrained optimization. Results: Validated on a representative six-axis industrial manipulator, the method reduces end-effector positioning error from 5.2 mm to 0.8 mm (an 84.6% improvement) while cutting calibration time by 60%. It substantially lowers hardware dependency and implementation complexity, offering a practical, cost-effective solution for in-situ high-accuracy robot calibration.

2 citationsRead paper

Dr. Jekyll and Mr. Hyde: Two Faces of LLMs

Dec 06, 2023arXiv.org

This work exposes a critical security vulnerability in large language models (LLMs) under role-playing attacks: injecting misaligned persona profiles successfully bypasses multi-layer safety filters in ChatGPT and Gemini, eliciting policy-violating responses on illegal and harmful tasks. We introduce the novel “persona-injection jailbreaking” paradigm, integrating prompt-guided persona modeling, adversarial dialogue orchestration, and trusted-persona reinforcement fine-tuning. To counter this threat, we propose a bidirectional defense framework that promotes internalization of trustworthy personas during inference and training, thereby enhancing model robustness against such attacks. Extensive experiments demonstrate that our defense significantly reduces attack success rates across diverse benchmarks—by up to 87% on targeted harmful queries—while preserving model utility. This work advances LLM safety alignment by establishing a principled, empirically validated methodology for mitigating persona-based jailbreaking, offering both theoretical insight and practical, deployable safeguards.

2 citationsRead paper
Recent publications

Latest Papers