Start Date
16-8-2018 12:00 AM
Description
In this article we are analyzing how Industrial Internet of Things (IIoT) sensors and devices behave while they undergo an attack. Using data generated from a controlled experiment where attacks were carried out on a Secure Water Treatment (SWaT) system, we analyze the behavior of the sensors. We observe that the readings from the sensors are non-linear in nature and resemble ECG waveform output, which helps in identifying inconsistencies or anomalies in heartbeats of patients. Through the comparison of sensor behavior during an attack and under normal conditions we find a significant difference in the features of the waveforms. Also, we look at the contrasting behavior of sensors under two different kinds of attacks: physical and cyber. The findings of this research motivate an alternative approach for anomaly detection and real time assessment of cyber-attacks on IoT devices with the use of analytics.
Recommended Citation
Onuchowska, Agnieszka; Chakraborty, Saurav; Jank, Wolfgang; and Shrivastava, Utkarsh, "Detection and Classification of Attacks on IoT Networks" (2018). AMCIS 2018 Proceedings. 24.
https://aisel.aisnet.org/amcis2018/AdvancesIS/Presentations/24
Detection and Classification of Attacks on IoT Networks
In this article we are analyzing how Industrial Internet of Things (IIoT) sensors and devices behave while they undergo an attack. Using data generated from a controlled experiment where attacks were carried out on a Secure Water Treatment (SWaT) system, we analyze the behavior of the sensors. We observe that the readings from the sensors are non-linear in nature and resemble ECG waveform output, which helps in identifying inconsistencies or anomalies in heartbeats of patients. Through the comparison of sensor behavior during an attack and under normal conditions we find a significant difference in the features of the waveforms. Also, we look at the contrasting behavior of sensors under two different kinds of attacks: physical and cyber. The findings of this research motivate an alternative approach for anomaly detection and real time assessment of cyber-attacks on IoT devices with the use of analytics.