Successive direct load altering attack in smart grid

摘要

In smart grid, a malicious entity can launch a direct load altering attack by injecting false commands into aggregators responsible for direct load control. It may remotely manipulate load, causing deviation in the operating frequency, and consequently lead to disruption in the system. In this paper, we mainly focus on the successive direct load altering attack, with which the attacker can continuously manipulate aggregators to achieve the larger impact. In addition to resulting in a larger impact, it is difficult for the controllers to detect such attacks as the attackers can inject false data to contaminate feedback data from aggregators to controllers. We present an attack model, and our analysis in this paper is from an attacker's perspective. Our model and analysis can serve as an important component also in the future for designing the counter strategies to such attacks. We propose a new frequency response model, which shows changes of the frequency undergoing a successive direct load altering attack. Attackers can utilize this model to analyze the impact of an attack sequence. Considering that attack sequences with different false commands can result in different levels of impact, we develop a three-step optimization method to analyze and find the optimal attack sequence. Our simulation results validate the feasibility and effectiveness of the successive direct load altering attacks. (C) 2018 Elsevier Ltd. All rights reserved.