Poisoning Attack on Load Forecasting

Liang, Yi; He, Di; Chen, Deming

doi:10.1109/isgt-asia.2019.8881664

Cited by 17 publications

(8 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…D val represents an untrained identically distributed validation set, as attackers are restricted to have no access to the test set, the attack effect is evaluated by model performance on D val . As the output at a single timestamp lacks specific meaning in regression problems, the current works design the attack effect W with overall performance, such as maximizing the MSE (Mean Square Error) of the output [22], [21].…”

Section: Existing Attack Methodsmentioning

confidence: 99%

“…Overall, for both data poisoning and adversarial attacks, most of the existing methods in the literature only aim to degrade the overall forecasting performance (e.g., maximizing the Mean Square Error (MSE) [22]). However, significantly decreased forecasting performance can be easily aware by the system operator or detected via deployed detection mechanisms.…”

Section: Existing Attack Methodsmentioning

confidence: 99%

“…Therefore, considering the feasibility of the attack, we assume the attacker can only tamper with variables in x forecast t for a data integrity attack. Note that attackers strictly follow attack limits, keeping ground truth y clean during data poisoning in the training phase, which is different from current poisoning attacks, such as [21], [22]. Further, in order to avoid excessive data tampering being detected by the system operator, the following constraint is enforced to the disturbance added by the attacker:…”

Section: Attackers' Knowledge and Capabilitymentioning

confidence: 99%

See 2 more Smart Citations

Vulnerability and Impact of Machine Learning-Based Inertia Forecasting Under Cost-Oriented Data Integrity Attack

Chen

Sun

Chu

et al. 2023

IEEE Trans. Smart Grid

View full text Add to dashboard Cite

With the increasing penetration of renewables, the power system is facing unprecedented challenges of low-inertia levels. The inherent ability of the system to defense disturbance and power imbalance through inertia response is degraded, and thus, system operators need to make faster and more efficient scheduling operations. As one of the most promising solutions, machine learning (ML) methods have been investigated and employed to realize effective inertia forecasting with considerable accuracy. Nevertheless, it is yet to understand its vulnerability with the growing threat of cyberattacks. To this end, this paper proposes a methodological framework to explore the vulnerability of ML-based inertia forecasting models, with a special focus on data integrity attacks. In particular, a cost-oriented false data injection attack is proposed, for the first time, with the primary objective to significantly increase the system operation cost while retaining the stealthiness of the attack via minimizing the differences between the pre-perturbed and after-perturbed inertia forecasts. Moreover, we propose four vulnerability assessment metrics for the ML-based inertia forecasting models. Case studies on the GB power system demonstrate the vulnerability and impact of the ML-based inertia forecasting models, as well as the stealthiness and transferability of the proposed cost-oriented data integrity attacks.

show abstract

Section: Existing Attack Methodsmentioning

confidence: 99%

Section: Existing Attack Methodsmentioning

confidence: 99%

Section: Attackers' Knowledge and Capabilitymentioning

confidence: 99%

See 1 more Smart Citation

Vulnerability and Impact of Machine Learning-Based Inertia Forecasting Under Cost-Oriented Data Integrity Attack

Chen

Sun

Chu

et al. 2023

IEEE Trans. Smart Grid

View full text Add to dashboard Cite

show abstract

“…GPS synchronized devices may be subject to well-known GPS spoofing attacks. In [116] authors describe poisoning attack on load forecasting mechanism in Smart Grid, while authors in [117] give brief overview of attacks on load forecasting.…”

Section: An Example Dosmentioning

confidence: 99%

A Survey of Denial-of-Service Attacks and Solutions in the Smart Grid

et al. 2020

View full text Add to dashboard Cite

The scope, scale, and intensity of real, as well as potential attacks, on the Smart Grid have been increasing and thus gaining more attention. An important component of Smart Grid cybersecurity efforts addresses the availability and access to the power and related information and communications infrastructures. We overload the term, Denial-of-Service (DoS), to refer to these attacks in the Smart Grid. In this paper, we provide a holistic and methodical presentation of the DoS attack taxonomies as well as a survey of potential solution techniques to help draw a more concerted and coordinated research into this area, lack of which may have profound consequences. To the best of our knowledge, the literature does not have such a comprehensive survey study of the DoS attacks and solutions for the Smart Grid. INDEX TERMS Denial-of-Service attacks, Smart grid security, Cybersecurity I. ACRONYMS In order to make the paper more clear and easier to read, we provide a table of acronyms in Table 1

show abstract

“…The studies in [28]- [30] have all focused on evasion attacks. To the best of our knowledge, the only study on poisoning attacks in power systems is done in [31], where the impact of poisoning attacks against short term load forecasting is studied.…”

Section: A Literature Reviewmentioning

confidence: 99%

Poisoning Attack against Event Classification in Distribution Synchrophasor Measurements

Kamal

Shahsavari

Mohsenian‐Rad

2021

2021 IEEE International Conference on Communications, Control, and Computing Technologies for Smart Grids (SmartGridComm)

View full text Add to dashboard Cite

Distribution-level phasor measurement units (D-PMUs), a.k.a., micro-PMUs, have received a growing attention in recent years to support various applications in power distribution systems. Many of the applications of micro-PMUs work based on the analysis of events in the stream of synchrophasor measurements to achieve situational awareness. A key step in almost every eventbased method in this emerging field is to classify the type of the event, where classification can be done with respect to various factors. However, if the task of event classification is compromised, then an adversary can highly affect the perception of the utility operator and undermine any event-based application that makes use of the event classification results. In this paper, we explore a new cyber-threat against data-driven event classification in micro-PMU measurements. In particular, we model the poisoning attack against support vector machine (SVM) as the method of event classification; which has been used in practice to study distribution synchrophasors. We apply the new attack model to an event classifier that uses real-world micro-PMU data. In addition to conducting vulnerability analysis, we also propose a novel attack detection method which can detect and evaluate the changes in the decision boundary of the SVM due to the poisoning attack. The proposed attack detection method is also able to identify the number of poisoned data points in the training dataset.

show abstract

Poisoning Attack on Load Forecasting

Cited by 17 publications

References 18 publications

Vulnerability and Impact of Machine Learning-Based Inertia Forecasting Under Cost-Oriented Data Integrity Attack

Vulnerability and Impact of Machine Learning-Based Inertia Forecasting Under Cost-Oriented Data Integrity Attack

A Survey of Denial-of-Service Attacks and Solutions in the Smart Grid

Poisoning Attack against Event Classification in Distribution Synchrophasor Measurements

Contact Info

Product

Resources

About