PhD Project Proposal - Trustworthy Machine Learning for Industrial Control Systems Security

Principal Supervisor(s): Dr Ihsen Alouani

Contact details: i.alouani@qub.ac.uk                                                                                           

Second Supervisor: Dr Kieran McLaughlin

Contact details: kieran.mclaughlin@qub.ac.uk

Proposed Project Title: Trustworthy Machine Learning for Industrial Control Systems Security

Project Introduction:

A recent study conducted by OpinionWay for the Club of Information and Digital Security Experts (CESIN), 57% of companies were victims of a cyberattack in 2020. And more than 8 out of 10 companies plan to acquire new technical solutions for cybersecurity and more than half want to increase the budget allocated to protection against cyber risks. With the deployment of Cyber-physical Systems in industrial environments, the importance of cybersecurity is further accentuated since the impact of an attack can go beyond the Cyber world and reach the physical world.

Much of cybersecurity is maintained by human expertise with a very high associated cost. To monitor the security of industrial control systems (ICS), event logs are mainly used. These logs are produced massively every day by all types of machines, applications, services, network equipment, etc. They constitute a voluminous quantity of data that must be analysed in real time to detect possible attacks, intrusions, and compromises.

Project Description:

Some anomaly detection processes within ICS require human expertise, which represents a great challenge, especially in terms of scalability of these security processes.

Machine Learning (ML) has established itself in recent years as an essential and effective tool in the automatic processing of large-volume data, particularly in tasks that require a certain expertise such as intrusion detection, anomalies or malicious activities recognition, etc. However, recent studies have shown that ML systems are vulnerable to adversarial noise; maliciously crafted perturbations to the input that force the models to output wrong labels.

In this project, we propose to investigate trustworthy ML systems for the security of ICS. Specifically, we propose to:

1. Design, test and deploy a machine learning approach for real-time threat detection in ICS.
2. Investigate the trustworthiness of the proposed approach against adaptive and sophisticated adversarial attacks.

iii. Propose defence mechanisms to harden our system against evasive and adaptive attackers.

Project Key Words: AI, Cybersecurity, Industrial Control Systems, Adversarial Attacks, Anomaly Detection

Full-Time                                               

Start Date: 1 October 2023

Application Closing date: 28 February 2023

Funding Body: DfE CAST

Project Funding Type: funded

Funding Information:  To be eligible for consideration for a DfE Studentship (covering tuition fees and maintenance stipend of approx. £17,668 per annum), a candidate must satisfy all the eligibility criteria based on nationality, residency and academic qualifications. The Studentship is open to UK and ROI nationals, and to EU nationals with settled status in the UK, subject to meeting the specific DfE nationality and residency criteria. Full eligibility information can be viewed via: https://www.economy-ni.gov.uk/publications/student-finance-postgraduate-studentships-terms-and-conditions  

This is an industrially sponsored project. An enhanced stipend (amount tbc) is payable to the sponsored student in addition to the stipend rate detailed above.

Academic Requirements: The minimum academic requirement for admission is normally an Upper Second Class Honours degree from a UK or ROI Higher Education provider in a relevant discipline, or an equivalent qualification acceptable to the University.

To Apply please complete an application through the Direct Applications Portal:

https://dap.qub.ac.uk/portal/user/u_login.php

Advert information