Early Detection of Distributed File-less Malware Using Vision Transformers (VC2644)

About the Project

University of the West of Scotland (UWS) is seeking to attract a PhD candidate of outstanding ability and commitment to join its vibrant and growing programme of internationally excellent research at its London Campus.

The rapid evolution of cyber threats has led to the emergence of highly sophisticated attack and evasion techniques such as file-less malware, which operates in system memory and leaves little or no trace on disk. These characteristics make such attacks particularly difficult to detect using both traditional and current security approaches.

This PhD project aims to address this critical and growing challenge by developing innovative, AI-powered methods for the early detection of distributed file-less malware. The project will explore the use of advanced deep learning techniques, with a particular focus on Vision Transformer (ViT) models, to analyse behavioural patterns derived from system activity. Unlike conventional approaches that rely on complete or centralised data, this research will investigate how early-stage and fragmented behavioural signals, collected across processes, systems, or time, can be effectively represented and utilised for detection.

A key aspect of the research involves transforming behavioural data, including system calls, memory usage, and network activity, into structured visual representations that can be analysed using transformer-based models. This enables the system to capture complex and long-range relationships within incomplete data, supporting early identification of malicious activity before significant system compromise occurs. The project will also adopt an experimental methodology, including the collection of behavioural data in controlled and secure environments (sandboxes), development of novel detection models, and rigorous evaluation against existing machine learning and deep learning approaches. Performance will be assessed in terms of detection accuracy, latency, and robustness under realistic, partially observable conditions.

This research aims to contribute to the advancement of proactive cyber security solutions, publish findings in leading academic venues, and provide valuable insights into the application of transformer-based AI models in complex, real-world threat scenarios.

The candidate/eligibility criteria

Essential Criteria:

1. A master’s degree in a relevant field, e.g. Cyber Security, Artificial Intelligence, Computer Science, or a closely related discipline, already obtained or near completion.
2. Demonstrable interest in cyber security and machine learning, and familiarity with research methods in a related area.
3. Knowledge of core cyber security concepts, including malware analysis, operating systems, and network security.
4. Experience with machine learning or deep learning frameworks, including TensorFlow and PyTorch.
5. Strong programming skills, particularly in Python, and familiarity with data analysis.
6. Strong analytical thinking and problem-solving abilities.
7. Strong written and verbal communication skills in English.
8. Eligibility for a UK-funded studentship (meeting residency requirements).

Desirable Criteria:

1. Experience working with large-scale, real-world malware datasets.
2. Proficiency in system-level programming languages (C/C++).
3. Prior research experience, including publications in the field.

The successful candidate must meet the following criteria:

• be a UK National (meeting residency requirements),
• or have settled status,
• or have pre-settled status (meeting residency requirements),
• or have indefinite leave to remain.

For more information, or to discuss the project informally, please contact Dr Danial Javaheri at Danial.Javaheri@uws.ac.uk

Application Deadline: 15/06/2026

Start Date: 01/10/2026

Applications must be made via the UWS Online System.

Funding Notes

This is a fully funded PhD Studentship and includes payment of tuition fees for 36 months at the home/UK rate and an annual maintenance stipend equivalent to UKRI minimum stipend rate (£21,805pa from 01/10/2026).

The successful candidate must meet the following criteria:

• be a UK National (meeting residency requirements),
• or have settled status,
• or have pre-settled status (meeting residency requirements),
• or have indefinite leave to remain