Static and Dynamic Load Altering Attack Detection Based on Tractable Optimization

Internet of Things (IoT)-enabled loads are gradually penetrating into the power grids. However, the vulnerabilities of IoT can be leveraged maliciously to compromise high-wattage appliances. This gives the adversaries a chance to launch load altering attacks(LAAs), which may result in system instability or even large-scale blackouts. Thus, detecting the LAAs in real time is a crucial task for grid operators. In this paper, we propose a novel detection method that can detect both static LAAs (SLAAs) and dynamic LAAs (DLAAs). In the end, we provide a tractable optimization-based reformulation that improves the scalability, and as such practicality, of the proposed method. This method can detect both static and dynamic LAAs in a unified frame. The experiments at last demonstrate the feasibility and effectiveness of the proposed detection method.