A botnet is a collection of internet-facing devices that are compromised and controlled by a malicious hacker. In this paper, we propose an attack utilising a botnet of high-wattage internet-facing devices, which we call a power botnet. Power botnet attacks can decrease the reliability of power supply, damage the power quality and even cause catastrophic consequences in power distribution grid. To study the effects on power distribution systems, we simulate three different types of power botnet attacks using OpenDSS, and show the change of OLTC lifespans under attacks. We then use deep learning methods to detect these attacks. We show successful detection for two of these attacks and a low detection rate for the third attack. To the best of our knowledge, this is the first paper to consider power botnet attacks, and leverage deep learning methods to detect these attacks on power distribution grids. Future work such as detection schemes for more complicated power botnet attacks will be developed based on the results of this work.
|Title of host publication||2019 IEEE Power and Energy Society General Meeting, PESGM 2019|
|Publisher||IEEE Computer Society|
|State||Published - Aug 2019|
|Event||2019 IEEE Power and Energy Society General Meeting, PESGM 2019 - Atlanta, United States|
Duration: 4 Aug 2019 → 8 Aug 2019
|Name||IEEE Power and Energy Society General Meeting|
|Conference||2019 IEEE Power and Energy Society General Meeting, PESGM 2019|
|Period||4/08/19 → 8/08/19|
Bibliographical noteFunding Information:
This work was supported in part by the National Science Foundation under Grant ECCS-1831811, in part by Eversource Energy, and in part by the Office of the Provost, University of Connecticut.
© 2019 IEEE.
- Attack Detection
- Cyber Security
- Load altering attack
- Machine Learning
- Power Botnet