Data-Driven False Data Injection Attacks Against Power Grids: A Random Matrix Approach

Abstract

We address the problem of constructing false data injection (FDI) attacks that can bypass the bad data detector (BDD) of a power grid. The attacker is assumed to have access to only power flow measurement data traces (collected over a limited period of time) and no other prior knowledge about the grid. Existing related algorithms are formulated under the assumption that the attacker has access to measurements collected over a long (asymptotically infinite) time period, which may not be realistic. We show that these approaches do not perform well when the attacker has a limited number of data samples only. We design an enhanced algorithm to construct FDI attack vectors in the face of limited measurements that can nevertheless bypass the BDD with high probability. The algorithm design is guided by results from random matrix theory. Furthermore, we characterize an important trade-off between the attack's BDD-bypass probability and its sparsity, which affects the spatial extent of the attack that must be achieved. Extensive simulations using data traces collected from the MATPOWER simulator and benchmark IEEE bus systems validate our findings.

Data Consistency for Data-Driven Smart Energy Assessment

“…The occurrence of an attack on PMUs can be identified through data-driven methods such as clustering algorithms (Wang X. et al, 2019 ) or by considering the correlation between the frequency adjustments implemented by the clock and the change in the measured phase angle (Shereen and Dán, 2020 ). Challenging types of attack act by knowing only data for a limited time period and are able to bypass the bad data detection procedures (Lakshminarayana et al, 2021 ). Cyber security applications protect devices, networks and data against digital attacks (Sun et al, 2018 ).…”

Quantity or Quality? Data Enabled Online Energy Dispatch

“…Wang et al propose a combined model-driven and datadriven method for online frequency stability assessment and control in [20]. Lakshminarayana et al design a data-driven algorithm for conducting false data injection to bypass bad data detector in [13]. Wu et al design a data-driven storage control framework facing dynamic prices in [21].…”

Learning Requirements for Stealth Attacks

“…with Λ denoting a random variable distributed as the AED of Λ in (51), and η is the abbreviation for the η-transform of Z T ΛZ that is solved from (24) for γ = 1.…”

“…It follows from Theorem 6 that we need to obtain the η-transform of Z T ΛZ from (24) to finalize the asymptotic characterization of the ergodic performance. The following proposition shows that (24) has a unique solution of η.…”

“…where the expectation is over the nonnegative random variable Λ, whose distribution is the AED of Λ defined in (51); and the value of η Z T ΛZ (1) is solved from (24), which is proved to be always solvable in Proposition 1.…”