Detecting Data Integrity Attacks in Smart Grid

Abstract

The smart grid, which is denoted as a next‐generation power grid system, integrates information and communication techniques in order to enable efficiency and reliability in energy service for consumers. This chapter shows a framework that can be used to explore different attack scenarios and design detection schemes to deal with data integrity attacks. It introduces the network model and the threat model, and then presents an approach to defend against data integrity attacks in the smart grid. The chapter then discusses the extension of work from the following aspects: detection schemes, scalability of detection, and other defensive schemes. It describes three detection schemes: statistical anomaly‐based detection, machine learning‐based detection, and sequential hypothesis testing‐based detection, and considers both line voltage and power load as detection features. The chapter also presents the performance evaluation conducted to validate the effectiveness of the investigated detection schemes to detect attacks. It explains the evaluation setup and then presents the results.