Autonomous Systems and Networks

The School of Engineering of the University of Glasgow is seeking a highly motivated graduates to undertake exciting 3.5-years PhD projects in the area of autonomous systems and networks. These projects would suit a computing scientist, or an engineer with interest in designing and implementing complex distributed systems. Candidates with non-traditional backgrounds are encouraged.

Success in these PhDs will not only address key challenges in the path to autonomous networks but will capitalize on strong industrial relationships for possible internships and collaboration.

Informal enquiries should be sent to Paul.Harvey.

Background information can be found here.

Topics

Telecommunication networks sit at the heart of our day-to-day lives and connect most of our professional, personal, and emergency activities. In recent years, these networks have become more like large, distributed data centers as hardware-based manually configured systems are replaced by software-virtualized programable infrastructure. Such flexibility and adaptability presents many interesting opportunities for network management and operation driven by software closed-loops. The destination of this journey is to be able to fully automate the operation and management of these networks with little or no involvement of engineers: autonomous networks.

To tackle this challenge, highly motivated graduates are sought to work on the below topic areas:

How to Validate Closed-Loops?

Testing these loops is challenging as it is impossible to create a replica of a full network and testing on live systems can lead to network blackouts.

Digital twins are digital replicas of a system that allow accurate testing and experimentation without effecting the live system. While there are many examples of digital twins for physical systems (e.g. planes or factory processes) there are fewer examples for distributed dynamic systems, as found in telecommunication networks.

PhD Goal: to explore the design and implementation of digital twins to validate software closed-loops for control and management in telecommunication networks. Questions to answer include: the degree of fidelity, scalability, reusability, and extensibility of digital twins of a highly distributed and dynamic system.

How to Apply Closed-Loops?

To ensure that different network operators and technology vendors have technology that is compatible and inter-operable, standards bodies (e.g. 3GPP) provide detailed standards to help technology development. Unfortunately, technology has gotten ahead of the standards meaning engineers focus on development of software and systems, rather than domain experts applying knowledge to solve problems.

In the programming languages and runtimes (PLR) community, creating inter-operable software in virtualised environments is a well-studied problem with many technologies having been developed and deployed. Exemplar languages and runtimes include Python, FORTH, Lua, Java, Java Script, the .NET framework.

PhD Goal: to explore the implications of portable language runtimes in the context of the distributed software-virtualized programable infrastructure found in telecommunication networks.