Intrusion Detection for Advanced Metering Infrastructure using Machine Learning

Advanced Metering Infrastructure (AMI) system, is made up of smart meters, data management software, and communication networks for energy usage analysis and data transmission. AMI technology has seen a surge in popularity all over the world due to its benefits over analogue systems. However, the system’s reliance on the internet, a public network, exposes them to vulnerabilities and cybersecurity threats. The inherent resource constraints of smart meter devices, such as limited memory and computing power, complicate the implementation of traditional security solutions, heightening concerns about cybersecurity risks. To address this challenge, we utilize machine learning techniques to develop a lightweight Intrusion Detection System (IDS) to protect AMI systems. In this study, we propose an Artificial Neural Network (ANN) model for the development of the Intrusion Detection System for AMI systems. Through rigorous testing on an AMI setup, our proposed IDS achieved remarkable success in detecting intrusion attacks of Denial of Service (DoS) and Distributed Denial of Service (DDoS), recording a 99.89% detection rate. The machine learning IDS proved efficient and lightweight, consuming less computing power and memory in its operations.