Analysis of False Data Injection Attacks on Multiarea Load Frequency Control

This paper analyses the False Data Injection Attack (FDIA) on tie-line sensor measurements of a Load Frequency Control (LFC) System from an attacker's perspective. The attack vector has been modeled considering an advanced LFC system, wherein, the measurements undergo State Estimation and bad data detection algorithms. Two different attack vector modeling scenarios have been considered: in the first model, an arbitrary attack vector modeling is performed to study its effect on the system operation; The later model is built systematically to induce sequential outages. The significant advantage of the proposed method is that the number of measurements to be attacked and system knowledge like topology data, line parameter values, transfer function parameters that are required is very low. The results are studied for an IEEE 39 bus, 3 area system.