Supervisors: Dr Chris Nemeth, Dr Simon Lunagomez-Coria and Dr Vasileios Giotsas
Start Date: October 2019 (flexible)
Incredible technological advances in data collection and storage have created a world in which we are constantly generating data. Analysing such large data sets creates significant statistical challenges, where it is necessary to trade between the speed and accuracy with which large volumes of data can be analysed and acted upon. Increasingly, recorded data exhibit a natural network structure, for example, friendship groups through social media. To extract key features from these data sets, it is important to account for the dependency in the network structure, however, including this dependence in our statistical models makes it more computationally challenging to analyse these data. The aim of this project is to develop statistically scalable inference tools for analysing large-scale network data, where the data may be temporally-evolving and what is learnt from the evolving network, and the associated uncertainty can aid in statistical learning procedures for identifying potential changes in network structure.
To stimulate this methodological development, the student will collaborate with Dr Giotsas on the problem of network traffic hijacking. Traffic hijacking attacks exploit the fact that the Internet's core routing protocol lacks built-in security mechanisms, allowing any adversary with access on the routing system to inject fraudulent routes and execute an array of attacks, including DDoS, impersonation and eavesdropping . We are now witnessing an alarming number of such traffic hijacking incidents that result in comprised critical infrastructure and services. The significant difficulty of preventing such attacks has attracted the attention of mainstream media , which illustrates the potential impact of this project.
The successful candidate will be required to spend up to two months each summer working at HIMR gaining relevant experience. To be considered for this studentship, candidates must be UK nationals and prepared to undergo security clearance procedures.
The studentship will be funded for a period of 4 years and covers the costs of university fees and will provide an annual tax-free stipend to the student (currently £16,776 per annum) plus increases for inflation.
Applicants are expected to have a masters level degree in a mathematical discipline containing a significant amount of statistics. The ideal candidate would be familiar with advanced statistical methods and possess strong computer programming skills.
Applicants interested in this position are encouraged to contact Dr Chris Nemeth (firstname.lastname@example.org) for informal enquires. To apply for the studentship, candidates should complete an online application naming Dr Chris Nemeth as the supervisor.
 Mitseva, Asya, Andriy Panchenko and Thomas Engel. "The state of affairs in BGP security: A survey of attacks and defences." Computer Communications (2018).
 Sherman, Justin. "It's far too easy for countries like Russia and China to hijack Internet traffic." Slate Magazine, Slate, 16 Nov 2018.