I. Introduction
The power grid is responsible for generation, transmission, and distribution of electricity to customers. To achieve these functions, the grid's control center (CC) performs certain auxiliary tasks to guarantee the quality of the grid's services and prevent hazards and disasters, e.g., blackouts. These tasks include monitoring the grid's status using local measurement units. Therefore, CC should assure the accuracy of these measurements using state estimator (SE). The traditional SE computes the difference between observed and expected measurements and compares the residual with a specific threshold. However, this technique is not realistic for the novel smart grid, as the power grid's exposure to communication networks leads to a new type of attack called false data injection (FDI) attack. FDI attack targets the grid infrastructure by injecting false measurements in the measurement units' periodic reports to the CC. The essential condition for successful FDI attack is to pass the SE test and avoid the alarm. The attacker forms the false injected measurements in a way that the residual value is still less than the assigned threshold. Accordingly, FDI attack can mislead CC to make wrong decisions that negatively impact on the grid's performance and consequently cause catastrophic results, such as large blackout. In addition, the FDI attacker may aim to gain financial profit by redistributing the power loads along the grid to illegally manipulate the electricity price and have cheaper power. For example, the adversary can buy the electricity by cheap price from a location in the grid, and then change the loads to cause a higher power price in another location to sell its electricity [1]–[5].