EECS Research Publications
Permanent URI for this collection
Browse
Recent Submissions
Item Re-identification and automatic update for biometrics(Springer Nature, 2025-01-01) Rattani, Ajita; Tistarelli, Massimo[No abstract available]Item Optimal placements for minimum GDOP with consideration on the elevations of access nodes(Institute of Electrical and Electronics Engineers Inc., 2024-11-18) Ding, Yanwu; Shen, Dan; Pham, Khanh D.; Chen, GensheMinimum geometric dilution of precision (GDOP) is desired for a high accuracy in the localization of an unknown node. The optimal placements for the access nodes (or sensing nodes) to achieve the minimum GDOP can be configured by combinations of symmetric cones or regular polyhedrons. The minimum GDOP occurs at the tips of the cones or the center of the polyhedrons. These points are referred to as the designated min-GDOP-points. To benefit the minimum GDOP, the unknown node needs to be at the min-GDOP-points. However, that is hardly the case in practice, because the unknown node could be anywhere in a 3-D working volume or a 2-D area. It is shown that the minimum value of GDOP is determined by the number of access nodes and the dimensions of the measurement, and it does not depend on the elevations of access nodes as long as the topology of the placement is not altered. This motivates the proposed design to use the elevations as an extra degree of freedom to improve the overall localization accuracy in the volume. The goal is to seek the optimal elevations of access nodes to reduce the averaged GDOP values over the volume while keeping the theoretical minimum GDOP value unchanged. The obtained solutions are verified by extensive simulations. © 2024 IEEE.Item Integrating function and topology for node criticality assessment in interdependent networks(IEEE Computer Society, 2024-12-03) Eslami, AliThe traditional topology-based metrics for assessing node criticality may fall short in highly heterogeneous multilayer networks. This work introduces a novel modeling framework that incorporates both function and topology to effectively measure node criticality. The framework is applied to the State of Kansas' transportation network as a practical example of an infrastructure network, identifying the most critical nodes under various operating load conditions. The findings reveal, among other insights, a significant rise in an adversary's success rate when targeting the most critical nodes rather than a random set of nodes. © 2024 IEEE.Item Adversarial machine learning for privacy-preserving space situational awareness(Institute of Electrical and Electronics Engineers Inc., 2024-11-26) Chauhan, Neha; Salinas Monroy, Sergio A.; Urien P.; Piramuthu S.Satellite operators rely on space situational awareness (SSA) data to avoid collisions with other space objects. However, SSA data can also be used by adversaries to launch both kinetic and non-kinetic attacks against satellites. An existing approach to prevent these attacks consists in labeling a sensitive satellite as a debris in the SSA data. Its objective is to hide the satellite's true label from potential adversaries. Unfortunately, adversaries can easily determine the true label of a satellite based on the orbit information provided by the SSA system. In this paper, we propose a novel adversarial machine learning approach to conceal sensitive satellites in SSA data from potential adversaries. Our method conceals the satellite by changing both its label and the orbit information. The resulting SSA data entry cannot be used to determine that the space object has been mislabeled. Our experimental results show that our proposed method reduces the ability of an adversary to determine the true label a of a concealed satellite to less than 50%. © 2024 IEEE.Item Automated detection of masquerade attacks with AI and decoy documents(Institute of Electrical and Electronics Engineers Inc., 2024-12-13) Berdychowski, Matt; Salinas Monroy, Sergio A.Adversaries can launch masquerade attacks by stealing user credentials and then logging into a system, pretending to be the compromised user. These attacks are particularly challenging to detect, because the adversaries have the same permissions as the users they impersonate. Previous works have shown that masquerade attacks can be detected by deploying decoy documents on the users' computers. Since decoy documents are, in general, only accessed by attackers, they can be used to detect masquerade attacks. However, previous works lack data generated by real-world adversaries and do not propose detection mechanisms. To address this challenge, we design effective machine learning masquerade attack detection models. The models are trained using data from our previous real-world experiment, which includes real cybercriminal interactions with a honeyfile server. We find that our machine learning models are able to detect masquerade attacks in a real-world scenario with an accuracy of at least 98%. © 2024 IEEE.