Coordinated Data-Injection Attack and Detection in the Smart Grid: A Detailed Look at Enriching Detection Solutions

Abstract

A smart grid improves the efficiency of power grids via the aid of modern communication, signal processing, and control technologies. While smart grid integration enables power grid networks to be smarter, it also increases the risk of cyberattacks due to the strong dependence on the cyberinfrastructure in the overall system. In this article, the coordinated datainjection attack detection problem in the smart grid is considered. Specifically, the data-injection attack model is first introduced and a thorough survey of existing detection methods is then given. Afterward, three important efforts to enrich the detection solution are presented in detail: 1) attacker versus defender dynamics, where possible interactive attack and defense strategies are discussed in the context of secure phasor measurement unit (PMU) placement 2) distributed attack detection and state recovery, where the focus is how to achieve the optimal centralized performance with a distributed approach 3) quickest detection (QD), where the trade off between the detection speed and detection performance is studied. A list of associated key open problems in this area is then presented to conclude this article.