Detection of False Data Injection Attacks: An Extending Dimension Data Modification Scheme

To achieve the detection of a typical stealthy false data injection attacks in two-channel, this paper presents a data modification scheme for extending the dimension of the measured value from the view of a defender. Before the sensory measurement data is transmitted to the communication network, the dimension of the data is expanded to change the system knowledge required by attacker to launch stealthy false data injection attacks. The dimension reducing processing of the measurement data transferred to the controller, causes a bias of the state estimation between attacker and defender to achieve the detection of attack. With the expanding-reducing processing of the data modification scheme, the defender can still detect residual discrepancies from the attacked measurement data and trigger an anomaly detector alarm, even though system knowledge modified by the data is obtained by the attacker. Finally, the effectiveness and realizability of the proposed data modification scheme are verified by numerical simulations.