Hybrid Approaches to Resilient Network Optimisation

Communication networks must remain operational in the face of numerous adversaries, including but not limited to natural disaster, power outage and targeted attack. Any algorithmic approach to the design, monitoring or control of such networks must seek to maximise performance in the presence of correlated failures. This gives rise to a family of nested, combinatorial problems which are likely to require a multilevel optimisation approach. 

This project involves the classification of such problems with relation to the polynomial time hierarchy and the creation of a general approach or framework to their solution which combines mathematical programming and metaheuristic techniques. Specific problems should be tackled in order to demonstrate the effectiveness of the approach. The successful completion of this project should give rise to numerous opportunities for publication.

The candidate will be expected to have experience of linear/integer programming techniques and will be required to produce hybrid solutions which combine these with metaheuristics. Experience of programming with industry standard solver APIs (such as Cplex or Gurobi) would be an advantage. A thorough knowledge of computational complexity, including the ability to prove membership of established complexity classes, is a requirement.

How to apply

Please click here for more information on how to apply.