Andrew Martin, Software Engineering Centre
David Wallom, Oxford e-Research Centre
MSc in Computer Science
Proposed Project
The idea of a computing grid has become a popular metaphor for describing the distribution of data and computation across a distributed collection of nodes. In Oxford we have a campus grid which involves over a thousand local computers, and also gives access to supercomputing resources locally and nationally.
A largely unsolved problem in grid computing is how to gain confidence that the results received from a grid job are genuine and secure: did some rogue system impersonate the genuine grid resources and provide false answers, or steal the data being processed? Did some legitimate system fall victim to a rogue administrator or a trojan?
The technologies of trusted computing offer the potential to improve this situation. We have some computers which contain the necessary hardware, and we are in the process of constructing a trusted grid testbed. This project will involve helping to make that happen. You help to connect the computers to the campus grid and work on the theory and practice of how to make their grid software aware of the trusted computing components. The main area of work will be making changes to the Condor system, to enable a cryptographically-strong identity for each node, and then to improve the trustworthiness of the central components.
This project would suit someone interested in distributed systems and security. Ideally you will already understand something about cryptography and protocols, and have some programming ability in C++.
For more details contact Andrew Martin and David Wallom.