Diagnosing False Data Injection Attacks in the Smart Grid: a Practical Framework for Home-area Networks

Advances in the metering infrastructure of the electric grid allow two-way communication capabilities between the utility center and a vast array of smart meters installed in the grid's distribution and transmission components. Nefarious users that manage to compromise insecure smart meters can alter the payload transmitted from these meters, and abruptly increase or reduce electricity demand in a coordinated manner. This malicious practice, known as false data injection attack, can destabilize the grid. This paper describes a practical framework for diagnosing false data injection attacks in the smart grid. We propose a behavioral-based monitoring system that can be installed at home-area networks for detecting the aforementioned anomalies. We demonstrate a real-world prototype of our system engineered with inexpensive devices such as Raspberry Pi's and Z-Wave wireless sensors, and evaluate its performance with real data.