Fully-funded PhD Studentships for specified projects in Computer Science

The Department of Computer Science at the University of Surrey is offering up to 12 fully-funded PhD studentships for specified projects (at UK rates) to strengthen its research. Studentships are available for the duration of 3.5 years (or 7 years at 50% time) for projects across the Department’s core research areas including cybersecurity and cryptography, distributed and concurrent systems, artificial intelligence and machine learning. Successful applicants will become part of a vibrant PhD community and will benefit from the strong research environment and high international visibility of the Department. 

Projects

The awarded studentships will be allocated to the PhD research projects listed below following a competitive selection process. Applicants are welcome to approach mentioned supervisors using the provided contact details to seek more information about the projects.

You may also be interested in our studentships on offer across all research areas in Computer Science.

Project 1: Provable security for tally-hiding elections

This project will focus on analysing security guarantees for tally-hiding electronic voting systems and designing cryptographic primitives for those systems. A secondary aim is to certify the security of these systems using formal-analysis tools. Alternative research directions (but not limited to) are on applied cryptography, provable security, privacy-preserving technologies, e-voting, and formal verification.

Contact: Dr. Catalin Dragan

Project 2: Natural language processing for longitudinal social and biomedical science datasets

This project will focus on metadata extraction and uplift for longitudinal text, such as social and biomedical studies from the UK Data Archive (UKDA). The student will apply deep learning techniques, including graph neural networks, for automated extraction of various data/metadata elements from non-uniformly structured text, as well as investigate mechanisms for language processing tasks such as semantic equivalence of text and text summarisation. The project will be in collaboration with social scientists and Natural Language Processing researchers from UCL, with scope for extension to multilingual text in collaboration with Tampere University (Tampereen korkeakoulusaatio sr)/Finnish Social Science Data Archive.

Contact: Dr. Suparna De

Project 3: Towards explainable machine learning: a fitness landscape approach

Multiple well-established algorithms to train neural networks there exist. However, as Yann LeCun observed, we know/understand very little about loss functions and their properties such as the behaviour of critical points. This doctoral project aims at better understanding the behaviour of loss functions by applying techniques used for characterising optimisation problems, the set of these techniques being known as Fitness Landscape Analysis (FLA). For the first time, the FLA result will be used to inform the design of novel training algorithms for neural networks. This project, albeit theoretical, has an impact at two levels: 1) it will enable to better interpret and explain the functioning of some machine learning algorithms; 2) it will enhance the performance on real-world tasks that are currently performed by machine learning trained by traditional gradient-based algorithms. A case study in Computer Vision will be used to exemplify the real-world impact of the proposed approach.   

Contact: Prof. Ferrante Neri

Project 4: Security and privacy for digital natives

This project aims to develop tools and theories to analyse the security and privacy of protocols that straddle the boundary between physical and digital systems. Examples are proving your physical identity in cyberspace, proving your vaccination status in real life, managing your NFTs, and contract signing by clicking on a button. You will work with and build formal models of such systems, define appropriate security and privacy goals and analyse whether these systems meet your goals.

Contact: Dr. Sasa Radomirovic

Project 5: A privacy algebra for personal data

Privacy is a major concern for individuals, organisations, and regulatory authorities in the Uk and globally. Clearly, understanding privacy implications is vital for empowering individuals to control their data and make informed decisions; however, there is a lack of expressive privacy notions (say privacy is partially or highly preserved) required to quantify privacy, particularly for heterogeneous data (say emails, documents, and photos) as existing research approaches in the field lack foundations for allowing individuals to specify desired privacy conditions over their personal data. In this project, we will design a privacy algebra for personal data by formalising privacy notions and defining new privacy measures for heterogeneous data, which eventually could lead to developing a suite of usable privacy tools.

Contact: Dr. Muhammad Rizwan Asghar

Project 6: Virtual-datacenter deployment and integration platform for optimal internet service deployment (VDDIP)

In this project, we aim to develop the architecture and prototype implementation of a multi-cloud “Virtual Data centers (VDCs) Deployment and Integration Platform” (VDDIP). VDDIP is a converged application fabric comprising of different hierarchical layers of controllers to deploy and manage VDCs automatically and optimally for the Application Service Providers (ASPs) & and Internet Service Providers (ISPs). Our proposed converged fabric handles various problems such as automated optimal service deployment across multiple clouds, forming optimal service function chains, resource allocation, finding optimal service paths, end-to-end monitoring and others.

Contact: Dr. Deval Bhamare

Project 7: Designing secure and fine-grained blockchain rewritings

Fine-grained blockchain rewritings allow anyone to create a mutable transaction associated with an access policy, and a modifier with sufficient rewriting privileges can modify the mutable transaction if her rewriting privileges satisfy the transaction's policy. This project aims to propose new cryptographic primitives to secure mutability.

Contact: Dr. Yangguang Tian

Project 8: Distributed and parallel graph algorithms

Distributed and parallel graph algorithms underpin many approaches for solving problems on large datasets or networks. This project will involve studying connections between distributed and parallel algorithms, and devising, analysing, and potentially implementing new algorithms for graph-based problems.

Contact: Dr. Peter Davies

Project 9. Privacy preserving signatures

Privacy preserving signatures constitute basic cryptographic ingredients for real world applications such as anonymous credentials, e-voting, and cryptocurrencies. This project will design state-of-the-art signatures focusing on more efficient constructions, stronger security guarantees, and additional properties (e.g, traceability, threshold, delegation).  Depending on the candidate’s interest, the project can have a post-quantum focus.

Contact: Dr. Dan Gardham

Project 10. Opacity of real-time systems: verification and enforcement

Opacity is a confidentiality property which means if a partially-observed system can prevent its visit to secret from being leaked. This property has been widely used to describe diverse scenarios in cyber-security/privacy problems, and its theoretical problems in untimed systems have been thoroughly investigated in the past two decades. This project will explore the opacity verification and enforcement problems in real-time systems, in order to enrich the theoretical and applicable studies in this area.

Contact: Dr. Kuize Zhang

Project 11. Computational modelling of neural network self-organization

Artificial neural networks, for instance deep neural networks, have revolutionized the field of Artificial Intelligence. However, current artificial neural networks are still very different from actual, biological neural networks. Here, the student will use innovative computational methods to simulate how biologically plausible neural networks develop in 3D and self-organize based on genetic rules.

Contact: Dr. Roman Bauer

Project 12. Joint 3D pose estimation of body, hand and face for the metaverse

The perception of users' poses is the key to many applications in a metaverse. This project aims to achieve fine-grained 3D pose estimation (reconstruction) of human body, hand and pose jointly. A new universal 3D model will be created and the cutting-edge deep learning models will be explored to achieve this ambitious goal.

Contact: Dr. Zhenhua Feng

Open to UK nationals, those with EU settled or pre-settled status, or indefinite leave to remain

We expect successful applicants to hold a BSc degree (with at least UK 2:1 honours, or equivalent) or an MSc degree with distinction in Computer Science or a related discipline.

The standard English language requirement is for a score of 6.5 or above (or equivalent) with 6.0 in each individual category, in an IELTLS Academic test taken in the last 2 years. Equivalent qualifications are listed on our language requirements page.