arXiv subjects

Zhuohuan Li

Publications and source records attributed to Zhuohuan Li.

3 recordsLinked to original sources

arXiv2026

Specification-Driven Generation and Evaluation of Discrete-Event World Models via the DEVS Formalism

World models are central to LLM agents that must evaluate actions over long horizons. Yet much existing work focuses on environments governed by physical dynamics or spatial structure, whereas many high-impact domains, including supply chains, procurement networks, and business processes, evolve through discrete events, timing constraints, and causal dependencies. These settings call for discrete-event world models. Existing approaches to constructing world models often fall near two extremes: hand-engineered simulators provide consistency and reproducibility, but are costly to build and adapt; neural models are flexible, but can suffer from compounding inconsistency over long-horizon rollouts. We seek a principled middle ground by synthesizing discrete-event world models online from natural-language specifications, retaining the reliability of explicit simulators while gaining the adaptability of neural models. We adopt the DEVS formalism and introduce a staged LLM-based generation pipeline that separates structural inference over component interactions from component-level event and timing logic. For evaluation, we develop benchmark suites in which simulators emit structured event traces, which are then validated against specification-derived temporal, causal, and semantic constraints. This enables reproducible verification and localized diagnostics. Together, these contributions produce world models that remain consistent over long-horizon rollouts, can be verified from observable behavior, and can be synthesized efficiently on demand during online execution.

cs.AI↗

arXiv2026

Energy Efficiency Testing and Modeling of a Commercial O-RAN System

Network energy efficiency is of critical importance to mobile network operators for economic and ecological reasons. The advent of the O-RAN architecture has brought disaggregation and virtualization, and in order to achieve the highest energy savings gains, we need rigorous measurement, analysis, and modeling of energy consumption at both the component and system levels. However, there remains a lack of publicly-available, quantitative data characterizing the behavior of commercial-grade O-RAN systems. In this white paper, we present a detailed energy-efficiency characterization and modeling of a commercial O-RAN system based on comprehensive power and performance measurements, using a network deployment that faithfully replicates a production O-RAN network deployed by a wireless carrier. The results are drawn from an energy test campaign conducted through a joint collaboration between the Open RAN Center for Integration and Deployment (ORCID) Lab Testing and Evaluation (T&E) Project and the Open Networking Foundation / Rutgers WINLAB Energy Efficiency R&D project. The test environment includes an O-RAN system with an AWS-hosted O-CU, a dedicated-server O-DU, and six high-power, multi-band O-RUs. Our results identify the dominant factors influencing power consumption across the O-RAN stack and quantify energy usage variation under different operational and traffic scenarios. These measurements can be used by operators to parameterize power-consumption models, ultimately supporting data-driven energy optimization and more sustainable operation of commercial O-RAN networks.

cs.NI↗

arXiv2023

Why Don't You Clean Your Glasses? Perception Attacks with Dynamic Optical Perturbations

Camera-based autonomous systems that emulate human perception are increasingly being integrated into safety-critical platforms. Consequently, an established body of literature has emerged that explores adversarial attacks targeting the underlying machine learning models. Adapting adversarial attacks to the physical world is desirable for the attacker, as this removes the need to compromise digital systems. However, the real world poses challenges related to the "survivability" of adversarial manipulations given environmental noise in perception pipelines and the dynamicity of autonomous systems. In this paper, we take a sensor-first approach. We present EvilEye, a man-in-the-middle perception attack that leverages transparent displays to generate dynamic physical adversarial examples. EvilEye exploits the camera's optics to induce misclassifications under a variety of illumination conditions. To generate dynamic perturbations, we formalize the projection of a digital attack into the physical domain by modeling the transformation function of the captured image through the optical pipeline. Our extensive experiments show that EvilEye's generated adversarial perturbations are much more robust across varying environmental light conditions relative to existing physical perturbation frameworks, achieving a high attack success rate (ASR) while bypassing state-of-the-art physical adversarial detection frameworks. We demonstrate that the dynamic nature of EvilEye enables attackers to adapt adversarial examples across a variety of objects with a significantly higher ASR compared to state-of-the-art physical world attack frameworks. Finally, we discuss mitigation strategies against the EvilEye attack.

cs.CR↗