An Anomalous Behavior Detection Method for IoT Devices Based on Power Waveform Shapes

In recent years, with the wide spread of the Internet of Things (IoT) devices, security issues for hardware devices have been increasing, where detecting their anomalous behaviors becomes quite important. One of the effective methods for detecting anomalous behaviors of IoT devices is to utilize operation duration time and consumed energy extracted from their power waveforms. However, the existing methods do not consider the shape of time-series data and cannot distinguish between power waveforms with similar duration time and consumed energy but different shapes. In this paper, we propose a method for detecting anomalous behaviors based on the shape of time-series data by incorporating a shape-based distance (SBD) measure. The proposed method firstly obtains the entire power waveform of the target IoT device and extract several application power waveforms. After that, we give the invariances to them and we can effectively obtain the SBD between every two application power waveforms. Based on the SBD values, the local outlier factor (LOF) method can finally distinguish between normal application behaviors and anomalous application behaviors. Experimental results demonstrate that the proposed method successfully detects the anomalous application behaviors, while the existing method fails to detect them.