Current Projects

Cyber-physical systems (CPS) integrate computation, networking, and physical processes, forming the backbone of modern infrastructure such as smart grids, autonomous vehicles, industrial control systems, and intelligent buildings. Their tight coupling of digital intelligence with physical operations makes them highly efficient and transformative, but also exposes them to unique security risks. A successful attack on a CPS can disrupt critical services, compromise safety, and cause far-reaching economic or societal damage. As these systems continue to expand in scale and complexity, rigorous security assessment and threat modeling become essential to ensure their resilience against emerging cyber and physical vulnerabilities.

This project extends the capabilities of Dante to explore the emerging landscape of AI-driven cyber-physical attacks. By leveraging Large Language Models (LLMs), we aim to autonomously generate novel attack strategies that exploit vulnerabilities in smart physical security systems. These AI-generated attacks are then injected into sensor datasets to rigorously test and evaluate the resilience of simulated sites. Through this approach, the project provides valuable insights into potential threats while advancing methods to strengthen the security of modern cyber-physical environments.

In light of the growing adoption of the Internet of Things (IoT), a new era of connectivity and convenience has emerged, fundamentally transforming various aspects of our daily lives. However, this increased interconnectivity also introduces significant security challenges, with Hardware Trojans standing out as a prominent concern. These malicious implants have the potential to compromise the integrity and reliability of IoT devices, posing threats to critical systems and sensitive data. It is imperative to address this pressing issue to ensure the ongoing reliability and trustworthiness of IoT ecosystems.

To tackle this challenge, our team of researchers at the Intelligent Cyber Systems and Architectures Laboratory (ICSA Lab) at North Carolina A&T State University is at the forefront of developing innovative methods for countering Hardware Trojans in IoT systems. Our latest contribution is the creation of the "IoT Playground," an open-source machine learning framework designed for evaluating the security of IoT devices. This framework provides a collaborative platform where the IoT community can collectively work together to address vulnerabilities related to Hardware Trojans.

Due to market capital, non-academic cyber-infrastructure is typically more up-to-date and provides higher security levels against cyber-attacks than academic cyber-infrastructure. In contrast, cyber-infrastructure used in academia often relies on legacy systems, which are particularly more vulnerable to attacks. In general, academic cyber-infrastructure differs from its non-academic counterpart in terms of data, workflows, users, and security concerns that should be considered while developing security countermeasures. 

In this project, we address the security concerns of academic cyber-infrastructures by modeling scientific cyber-infrastructure using the concept of a cybersecurity knowledge graph (CSKG) that encodes vulnerabilities, threats, and mitigation strategies in the form of a directed graph. The model is used to conduct a security assessment on a set of cyber- infrastructures within a coalition of higher education institutions located in the state of North Carolina. The model is tuned using the data collected from the member institutions and the data available in other databases, e.g., National Vulnerability Database (NVD).