Deep Latent Space Clustering for Detection of Stealthy False Data Injection Attacks Against AC State Estimation in Power Systems

Bhattacharjee, Arnab; Mondal, Arnab Kumar; Verma, Ashu; Mishra, Sukumar; Saha, Tapan Kumar

doi:10.1109/tsg.2022.3216625

Cited by 8 publications

(4 citation statements)

References 60 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Advantageously, this method requires only a few labelled data to train a well‐performing detector. To eliminate label limitations, [67] proposed a self‐supervised clustering model with a stacked AE network. It can achieve clean clustering of data into benign and compromised samples without labelled supervision.…”

Section: Review Of Deep Learning Applications For Cybersecurity In Sm...mentioning

confidence: 99%

“…Using deep learning as auxiliaries [52,66,72,74] Simply developing classifiers for detection [53,54,69,76,77,81] Locating false data injection attacks [55,56,60,63,68,75] Resorting to deep reinforcement learning for detection [64,65,79,95] Detecting attacks with specific targets [57,70,78,80] Addressing the problem of attack samples insufficiency [58,59,67,83] Considering disturbances from renewable energy integration [60,61] Handling the privacy problem in constructing detectors [62,71,73] [51] designed novel FDIA strategies by introducing adversarial samples (also called perturbation vectors) into FDIAs, thereby deceiving BDDs and DL-based detectors.…”

Section: Classification Literaturementioning

confidence: 99%

See 1 more Smart Citation

Deep learning for cybersecurity in smart grids: Review and perspectives

Ruan

Liang

Zhao

et al. 2023

Energy Conversion and Econom

View full text Add to dashboard Cite

Protecting cybersecurity is a non‐negotiable task for smart grids (SG) and has garnered significant attention in recent years. The application of artificial intelligence (AI), particularly deep learning (DL), holds great promise for enhancing the cybersecurity of SG. Nevertheless, previous surveys and review articles have failed to comprehensively investigate the intersection between DL and SG cybersecurity. To address this gap, this study presents a survey of the latest advancements in DL technology and their relevance to SG cybersecurity. First, the functional mechanisms and scope of application of common DL techniques are explored. Subsequently, SG cyberthreats are categorised into distinct types of cyberattacks that have not been systematically examined in previous surveys. Based on this, a thorough review of the application of DL techniques in addressing each cyberthreat along with recommendations and a generalised framework for enhancing cyberattack detection using DL is offered. Finally, insights are provided into the emerging challenges presented by DL applications in SG cybersecurity that are yet to be widely acknowledged, and potential research avenues are proposed to address or alleviate these challenges.

show abstract

Section: Review Of Deep Learning Applications For Cybersecurity In Sm...mentioning

confidence: 99%

Section: Classification Literaturementioning

confidence: 99%

Deep learning for cybersecurity in smart grids: Review and perspectives

Ruan

Liang

Zhao

et al. 2023

Energy Conversion and Econom

View full text Add to dashboard Cite

show abstract

“…ΨALG ∈ R m×m and ΘALG ∈ R m×m are positive definite diagonal matrices. Differentiating VALB in (31) with respect to time and using (29), one obtains ( )…”

Section: Alg-based Cooperative Strategymentioning

confidence: 99%

“…To avoid this, much effort should be devoted to conquering the detrimental consequence caused by FDI attacks. Recently, various control methods of electric systems under the FDI attacks have been developed [30][31][32][33], while the resilient formation control approaches of multiagent systems subject to FDI attacks on the control channels are relative rare. To alleviate the adversarial effect arises from FDI attacks, several robust controllers have been presented, one of these controllers is to employ sliding mode control technology to address the FDI attacks on the control channels (or actuators), which can rapidly enhance the convergence rate and achieve control objectives, and then maintain the desired performance of the actuators despite emergence of transient chattering.…”

mentioning

confidence: 99%

Adaptive Learning Gain based Resilient Control of Multiagent Systems with Partial Uncertain Control Coefficients Under FDI Attacks

Yang,

Long,

Lei

2023

Preprint

View full text Add to dashboard Cite

This paper presents a novel approach to security resilient control in collective multiagent systems, incorporating an adaptive learning gain mechanism. The proposed method addresses challenges posed by bounded external disturbances, partial false-data-injection (FDI) attacks on the actuators, and partial unknown time-varying control coefficients. The presented synthesized control consists of following parts. Specifically, a robust adaptive control law with implicit learning is proposed to address the modeling parameterized uncertainties and disturbances. An adaptive learning gain-based control technique is designed to reduce the adversarial effect caused by mismatched disturbances with time-varying parameters, which curbs the cooperative errors to the assigned field. A resilient control approach is established to mitigate the detrimental impact of malicious attacks on actuators. An adaptive control strategy, utilizing a novel congelation of variables, is implemented to address the control issue pertaining to partially unknown control coefficients. This technique ensures that control inputs align with the practical actuators by establishing feasible control bounds. The property of the resulting closed-loop error multiagent systems is analyzed. Finally, the efficacy of the designed controller is illustrated by providing an illustrative example.

show abstract