A Novel Distributed and Stealthy Attack on Active Distribution Networks and a Mitigation Strategy

Abstract

Rapid advances in smart devices tremendously facilitate our day-to-day lives. However, these can be exploited remotely via existing cyber vulnerabilities to cause disruption at the physical infrastructure level. In this paper, we discover a novel distributed and stealthy attack that uses malicious actuation of a large number of small-scale loads residing within a distribution network (DN). This attack is capable of cumulatively violating the underlying operational system limits, leading to widespread and prolonged disruptions. A key element of this attack is the efficient use of attack resources, planned via Stackelberg games. To mitigate this type of an attack, we propose a countermeasure strategy which adaptively suppresses adverse effects of the attack when detected in a timely manner. The effectiveness of the proposed mitigation strategy is demonstrated via theoretical convergence studies, practical evaluations and comparisons with the state-of-the-art using realistic load flow and DN infrastructure models.

Total Variation-Based Distributed Kalman Filtering for Resiliency Against Byzantines

“…5, in order to observe the fluctuation caused by the selection matrices, we plot the MSE in (38) and MSE ′ = 1 L lim k→∞ L i=1 tr(P i (k)) with P i (k) in (8). 1 Thus, we can examine the accuracy of our theoretical finding to compute the expected value of the error covariance with respect to the selection matrices in (23). The comparable performance of the MSE and MSE ′ in Fig.…”

“…with P i being the steady-state error covariance matrix of agent i in (23). The simulation results presented in the following are obtained by averaging over 100 independent experiments.…”

“…To this end, [21] designs a false data-injection strategy that maximizes the trace of error covariance in a remote state estimation scenario. An event-based stealthy attack strategy that degrades estimation accuracy is proposed in [22], while [23] develops a stealthy attack strategy and its optimal defense mechanism by exploiting power grid vulnerabilities. Furthermore, [24] proposes a linear stealthy attack strategy and its feasibility constraints.…”

Robustness assessment of the cyber‐physical system against cascading failure in a virtual power plant based on complex network theory

“…Studies on the vulnerability of the power grid structure were generally based on the complex network theory, especially the analysis and verification of small‐world (SW) characteristics and scale‐free characteristics in power systems 9,10,11,12,13 . However, some scholars focused on the vulnerability analysis of ADNs 5,6,14 . Cao et al 5 proposed a routing algorithm based on the link‐state protocol to determine the optimum data transmission paths in cyber networks.…”

“…Liu et al 6 proposed a cyber‐physical active distribution system to minimize the construction and operation cost of CPSs and maximize the penetration level of distributed generation. Srikantha et al 14 proposed a population game using real‐time and adaptive defense mechanisms based on data aggregation to mitigate the adverse effects of the Stackelberg game attacks that used malicious actuation of large number of small‐scale loads residing in a distribution network. Small‐scale and large‐scale test networks were simulated as a CPS in ADNs to verify the feasibility and effectiveness of a hybrid state estimator and hybrid state estimator based single‐phase‐to‐ground fault line identification methods 7 .…”