Detection of false data injection attacks in cyber-physical systems using graph convolutional network

Smart grids comprise numerous related units based on intelligent communication and information-oriented technologies. The advanced network communication technologies are the building blocks of these intelligent grids, allowing the real-time transmission of power state information. However, these systems are highly prone to cyberattacks, especially false data injection (FDI) attacks that disturb the normal management of power grids by injecting false state estimation values. This paper proposes a graph convolutional network (GCN) framework to detect FDI attacks. This technique analyses the graphical aspect of FDI attacks by exploiting the graphical structures of the power network to analyse the fluctuating state estimation values based on the system topology and detect the location of FDI attacks. The standard IEEE 30, 118 and 2848-bus systems are employed to evaluate the efficiency of the proposed technique. Simulation results show that the proposed approach can efficiently detect FDI attacks in small and large cyber-physical systems with reasonable accuracies and detection times, considering different magnitudes of disturbances and attack sparsities.