PhD Studentship in Computer Science: Autonomous Resilience in Heterogeneous IoT Networks

Award Summary

100% home fees covered, and minimum tax-free annual living allowance of £21,805 (2026-27 UKRI rates)

Overview

Modern infrastructure, from smart cities to industrial IoT (IIoT), relies on massive networks of heterogeneous devices operating in complex, dynamic environments. However, traditional fault tolerance mechanisms struggle to manage the runtime uncertainties and diverse failure modes inherent in these systems. This research pioneers the fusion of distributed computing and artificial intelligence (AI) to create a resilient, next-generation architecture.

The project aims to empower IoT systems with "self-diagnosis" and "self-healing" capabilities. By integrating federated learning, graph neural networks, and blockchain technology, we will develop a framework that moves beyond static maintenance toward an autonomous, trustworthy system capable of real-time root cause analysis and automated recovery.

The Research Challenges

Ensuring the resilience of distributed IoT systems involves overcoming significant technical hurdles:

a) Adaptive Fault Monitoring: Traditional systems fail to identify root causes in multi-modal data streams; this project utilizes federated learning and graph neural networks for real-time, adaptive diagnosis.

b) Uncertainty in Dynamic Environments: Runtime uncertainties require sophisticated risk modeling; we will employ Bayesian deep learning and deep reinforcement learning to quantify uncertainty and forecast potential faults.

c) Trustworthy Autonomous Recovery: Autonomous self-healing must be verifiable; this challenge is met by designing AI-driven decision engines integrated with blockchain to ensure all recovery operations are automated and credible.

d) Scaling Heterogeneous Networks: The sheer scale of modern IoT demands a theoretical framework that remains resilient despite the complexity of diverse device types and interconnections.

Number Of Awards

1

Start Date

21 September 2026

Award Duration

3.5 years

Application Closing Date

13 May 2026

Sponsor

School of Computing

Supervisors

Dr. Yinhao Li

Eligibility Criteria

You must have/expect to gain, minimum 2:1 Honours degree/international equivalent in a relevant subject to the proposed PhD project. A solid background in computer science or systems engineering. Knowledge of AI/ML algorithms, particularly graph neural networks and reinforcement learning, is highly advantageous. A keen interest in distributed computing, IoT architecture, and system resilience is essential to align with the core objectives of this research. Additionally, candidates should demonstrate an interest in the practical application of these technologies within smart cities and industrial IoT contexts.

The studentship covers fees at the Home rate (UK and EU applicants with pre-settled/settled status and meet the residency criteria). International applicants are welcome to apply but must cover the difference between Home and International fees.

Applicants whose first language is not English require IELTS score of 6.5 overall, minimum of 5.5 in all sub-skills.

International applicants may require ATAS clearance certificate prior to obtaining their visa.

How To Apply

Use NewcastlePortal 

Select ‘Create Postgraduate Application’.  

Use ‘Course Search’:  

• search ‘Course Title’ programme code:8050F
• Research Area: Computing Science
• Select ‘PhD Computer Science’ as programme

You will need to provide information in ‘Further Questions’ section:  

• ‘Personal Statement’ (mandatory) - upload document or write statement directly into form  
• studentship code COMP2176 in the ‘Studentship/Partnership Reference’ field 
• when prompted for how you are providing your research proposal - select ‘Write Proposal’. Type title of research project from this advert and provide your own statement.

Contact Details

Yinhao Li

Advert information