Stealth false data injection using independent component analysis in smart grid

Esmalifalak, Mohammad; Nguyen, Huy; Zheng, Rong; Han, Zhu

doi:10.1109/smartgridcomm.2011.6102326

Cited by 117 publications

(69 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The use of independent component analysis in [13] is the most relevant. The authors of [13] proposed to identify a mixing matrix from which to construct an unobservable attack. However, such techniques require that loads are statistically independent and non-Gaussian, and the techniques need full sensor observations.…”

Section: B Related Work and Organizationmentioning

confidence: 99%

“…The fusion center observes corrupted measurements instead of the real measurements . The adversarial modification is mathematically modeled by: (12) where is an attack vector, and is the set of feasible attack vectors defined as (13) Liu, Ning, and Reiter [6] presented an unobservable attack, which is a powerful attack mechanism capable of perturbing the state estimate without being detected. An unobservable attack can be formally defined as follows.…”

Section: Adversary Modelmentioning

confidence: 99%

See 1 more Smart Citation

Subspace Methods for Data Attack on State Estimation: A Data Driven Approach

Kim

Tong

Thomas

2015

IEEE Trans. Signal Process.

188

115

View full text Add to dashboard Cite

Data attacks on state estimation modify part of system measurements such that the tempered measurements cause incorrect system state estimates. Attack techniques proposed in the literature often require detailed knowledge of system parameters. Such information is difficult to acquire in practice. The subspace methods presented in this paper, on the other hand, learn the system operating subspace from measurements and launch attacks accordingly. Conditions for the existence of an unobservable subspace attack are obtained under the full and partial measurement models. Using the estimated system subspace, two attack strategies are presented. The first strategy aims to affect the system state directly by hiding the attack vector in the system subspace. The second strategy misleads the bad data detection mechanism so that data not under attack are removed. Performance of these attacks are evaluated using the IEEE 14-bus network and the IEEE 118-bus network.Index Terms-Cyber physical system, data framing attack, false data injection, state estimation, subspace method.

show abstract

Section: B Related Work and Organizationmentioning

confidence: 99%

Section: Adversary Modelmentioning

confidence: 99%

Subspace Methods for Data Attack on State Estimation: A Data Driven Approach

Kim

Tong

Thomas

2015

IEEE Trans. Signal Process.

188

115

View full text Add to dashboard Cite

show abstract

“…These works assume an omniscient attacker knowing the system parameters and topology. For more practical attack scenarios, data-driven attacks that learn the necessary system information from partial sensor observations are proposed [11,23,26]. Section 3 reviews the detail of the vulnerability analysis and the attack mechanisms for unobservable state attacks.…”

Section: Related Workmentioning

confidence: 99%

“…For a general matrix H, this problem has been known to be NP-complete (e.g., see Corollary 1 in [43].) However, Kosut et al show in [28] that the special structure of H can be exploited to have a polynomial time algorithm for solving (11). The solution of (11) provides the smallest S A that enables an unobservable state attack and the sparsest attack vector.…”

Section: Cost-effective Attack Strategymentioning

confidence: 99%

“…The sensor data are more accessible than the topology or the line parameter information, because they can be acquired by hacking into local communications links or devices. Esmalifalak et al in [11] made the first such attempt based on independent component analysis (ICA). They employ ICA to estimate a mixing matrix HA based on sensor measurements, where A is a linear operator that maps the load profile to the state vector.…”

Section: Data-driven Attack Mechanismmentioning

confidence: 99%

See 1 more Smart Citation

Against Data Attacks on Smart Grid Operations: Attack Mechanisms and Security Measures

Kim

Tong

2015

Cyber Physical Systems Approach to Smart Electric Power Grid

View full text Add to dashboard Cite

Abstract. This chapter provides a survey and some highlights of recent developments on cyber security issues related to smart grid operations. In particular, we present data attack models and attack mechanisms on system state estimation, generation dispatch, and market operations. Security measures via sensor protection and data authentication are discussed. Although presented in the context of a smart grid, the main ideas are applicable to general cyber physical systems.

show abstract