Authentication of cyber-physical systems under learning-based attacks

Khojasteh, Mohammad Javad; Khina, Anatoly; Franceschetti, Massimo; Javidi, Tara

doi:10.1016/j.ifacol.2019.12.183

Cited by 5 publications

(6 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The smart city architecture is highly prone to malicious attacks that include different injection attacks such as code injection attacks, SQL injection attacks, code-reuse attacks, etc. However, injection attacks normally inject malicious code into an application to disrupt its functioning, exploiting untrusted resources [ 20 , 21 , 22 ].…”

Section: Smart City Layered Adversariesmentioning

confidence: 99%

A Comparative Analysis on Blockchain versus Centralized Authentication Architectures for IoT-Enabled Smart Devices in Smart Cities: A Comprehensive Review, Recent Advances, and Future Research Directions

Khalil

Malik

Uddin

et al. 2022

Sensors

View full text Add to dashboard Cite

Smart devices have become an essential part of the architectures such as the Internet of Things (IoT), Cyber-Physical Systems (CPSs), and Internet of Everything (IoE). In contrast, these architectures constitute a system to realize the concept of smart cities and, ultimately, a smart planet. The adoption of these smart devices expands to different cyber-physical systems in smart city architecture, i.e., smart houses, smart healthcare, smart transportation, smart grid, smart agriculture, etc. The edge of the network connects these smart devices (sensors, aggregators, and actuators) that can operate in the physical environment and collects the data, which is further used to make an informed decision through actuation. Here, the security of these devices is immensely important, specifically from an authentication standpoint, as in the case of unauthenticated/malicious assets, the whole infrastructure would be at stake. We provide an updated review of authentication mechanisms by categorizing centralized and distributed architectures. We discuss the security issues regarding the authentication of these IoT-enabled smart devices. We evaluate and analyze the study of the proposed literature schemes that pose authentication challenges in terms of computational costs, communication overheads, and models applied to attain robustness. Hence, lightweight solutions in managing, maintaining, processing, and storing authentication data of IoT-enabled assets are an urgent need. From an integration perspective, cloud computing has provided strong support. In contrast, decentralized ledger technology, i.e., blockchain, light-weight cryptosystems, and Artificial Intelligence (AI)-based solutions, are the areas with much more to explore. Finally, we discuss the future research challenges, which will eventually help address the ambiguities for improvement.

show abstract

Section: Smart City Layered Adversariesmentioning

confidence: 99%

A Comparative Analysis on Blockchain versus Centralized Authentication Architectures for IoT-Enabled Smart Devices in Smart Cities: A Comprehensive Review, Recent Advances, and Future Research Directions

Khalil

Malik

Uddin

et al. 2022

Sensors

View full text Add to dashboard Cite

show abstract

“…When the controller detects an attack, it shuts the system down and prevents the attacker from causing further "damage" to the plant. The controller/detector is aware of the plant dynamics (1) and knows the open-loop gain a of the plant. On the other hand, the attacker knows the plant dynamics (1) as well as the plant state X k , and control input U k (or equivalently, Z k ) at time k (see Fig.…”

Section: Problem Setupmentioning

confidence: 99%

“…The controller/detector, being aware of the dynamic (1) and the open-loop gain a, attempts to detect possible attacks by testing for statistical deviations from the typical behavior of the system (1). More precisely, under the legitimate system operation (corresponding to the null hypothesis), the controller observation Y k behaves according to…”

Section: B Detectionmentioning

confidence: 99%

“…Finding the optimal privacyenhancing signal is left for future research. We remark that since the submission of this paper [1], [52], some latter literature has appeared that builds on it. In particular, in a follow-up work [60], Ziemann and Sandberg focused on designing optimal privacy-enhancing signal, by studying the optimal control problem of linear systems regularized with Fisher information, where the latter serves as a proxy to the estimation quality of A via the Cramer-Rao lower bound.…”

Section: Privacy-enhancing Signalmentioning

confidence: 99%

See 1 more Smart Citation

Learning-Based Attacks in Cyber-Physical Systems

Khojasteh

Khina

Franceschetti

et al. 2021

IEEE Trans. Control Netw. Syst.

Self Cite

View full text Add to dashboard Cite

We introduce the problem of learning-based attacks in a simple abstraction of cyber-physical systems-the case of a discrete-time, linear, time-invariant plant that may be subject to an attack that overrides the sensor readings and the controller actions. The attacker attempts to learn the dynamics of the plant and subsequently overrides the controller's actuation signal, to destroy the plant without being detected. The attacker can feed fictitious sensor readings to the controller using its estimate of the plant dynamics and mimic the legitimate plant operation. The controller, on the other hand, is constantly on the lookout for an attack; once the controller detects an attack, it immediately shuts the plant off. In the case of scalar plants, we derive an upper bound on the attacker's deception probability for any measurable control policy when the attacker uses an arbitrary learning algorithm to estimate the system dynamics. We then derive lower bounds for the attacker's deception probability for both scalar and vector plants by assuming an authentication test that inspects the empirical variance of the system disturbance. We also show how the controller can improve the security of the system by superimposing a carefully crafted privacy-enhancing signal on top of the "nominal control policy." Finally, for nonlinear scalar dynamics that belong to the Reproducing Kernel Hilbert Space (RKHS), we investigate the performance of attacks based on nonlinear Gaussian-processes (GP) learning algorithms.

show abstract

“…The problem of cyber-attacks in networked control systems has been studied comprehensively in previous research [4]- [8]. In this paper our focus is on the analysis and detection of replay attacks in control systems [9], [10].…”

Section: Introductionmentioning

confidence: 99%