Real-time implementation of IoT Enabled Cyber Attack Detection System (IoT-E-CADS) in Advanced Metering Infrastructure (AMI) using Machine Learning Technique (MLT)

Abstract

Abstract The introduction of Internet of Things (IoT) and Smart Energy Meters (SEM) in power grid expects high level cyber security for properly operating the system. This paper proposes an IoT Enabled Cyber Attack Detection System (IoT-E-CADS) in Advanced Metering Infrastructure (AMI) using Machine Learning Technique (MLT). The proposed Bi-level IoT-E-CADS is working in the industry standards for the detection of two types of attacks in the smart grid environment. In the first level, the Isolation Forest machine learning algorithm is used to find the cyber attacks and anomaly detection in real time system. In the second level, Decision Tree machine learning algorithm is used to find the cyber attacks and False Data Injection in a real time system. The designed hardware is implemented and tested at Quantanics Techserv Pvt. Ltd., located in Madurai, Tamil Nadu, India. This company has AMI facility with 10 smart meters, one data concentrator and dedicated server system. The company energy profile and all electrical parameters are monitored and stored using AMI facility. The proposed IoT-E-CADS successfully implemented in this location and detect the manually created two types of cyber attacks. Based on the obtained results, it is observed that the IoT-E-CADS is able to detect cyber threats with the accuracy level of 95% and provides a complete cyber security solutions for secured monitoring unit in commercial environment