Machine Learning based Cyber Threat Intelligence and Detection in the Smart Grid

点击翻译按钮获取中文摘要

Information and Communications Technologies (ICTs) integration with the electric power grid has evolved**into the Smart Grid. The aim is to incorporate internet integrated sensors, actuators, controllers and embedded**systems into electricity grids to make intelligent and better decisions in managing the customer demand and**optimizing available resources. However, ICT amalgam with grid has opened new avenues for cyber attacks.**To develop cybersecurity techniques is essential for the protection of the smart grid.**In consultation with our supporting organization, N-Dimension Solutions Inc. a leading Canadian cybersecurity**protection service provider for critical infrastructures. We have agreed to undertake machine learning research**that can improve their managed security service for utilities that combines an Intrusion Detection System (IDS)**with timely alerts, in-depth threat analytics, and actionable guidance on threat remediation and risk mitigation,**to help utilities protect their network, data and assets. Currently, N-Dimension solution lacks cognitive learning**abilities. The objective of this project is to model the behaviour of the smart grid's system. That will enable the**IDS in efficiently detecting cyber attacks utilizing the machine learning techniques developed in the result of**this research activity.**Machine learning based application is thought to be promising for cybersecurity threat detection and**intelligence, but its research is still in infancy. N-Dimension is looking to incorporating machine learning**techniques within its current platforms, in order to improve cyber attacks detection in the smart grid networks.**This project can help in improving cyber attacks detection for Canadian critical infrastructure operators. The**project will also provide essential training for the highly qualified personnel to join the Canadian high-tech**industry.