Research

My published papers, book chapters and journals from my university days can be found here.

While most of my research was embodied through telecommunications the under-current and my main interest is the behaviour of large systems. These can be man-made or natural systems, including, but not limited to:

• Telecommunications networks: mostly network evolution and co-adaptation (my PhD thesis topic) and as network planning, scalability, routing, resilience, self-healing and demand prediction..
• Transport networks: this is really a generalised case of the telecommunications network above - these networks include power distribution networks, road haulage networks. Please see Fully Meshed Networks for my current work in electrical power network planning, management, forecasting and dimensioning.
• Social networks: group behaviour, flocking, distributed decision processes, game theory, information propagation.. Drop by Fully Meshed Labs if you need any large networks analysed.
• Biological and natural systems: allometric scaling, the cardiovascular system, trees, rivers, the weather..
• Financial markets: this one is a little different to the others - for one the rules keep changing (changes in regulation, the creation of new financial instruments, new timescales etc.), and the other big difference is rationality, or the lack thereof. There are many traits in common with the others, though: information asymmetry, propagation delay, hidden couplings (e.g. bonus mechanisms, spheres of influence), influence from other "layers" (e.g. commodity production volumes).

.. and anything else that is big, interconnected and has interesting dynamic behaviour.

My research interest then, is generally broader than, but could approximately be described as:

Complexity and Emergence
Under certain circumstances a large interconnected system can adapt and perform computation which is not explicitly part of the elements of the system. I have worked on various aspects and abstractions of telecoms networks as dynamical systems and investigated their ability to organise and adapt, and examined the conditions under which this occurs (and related topics such as Catastrophe/Bifurcation theory). Under the software section of this site you can find two software packages I wrote related to investigating large scale system behaviour:

qBGPSim is an inter-domain network simulator that uses aggregate traffic models to estimate the effect of congestion on traffic flows and then feeds this back to the EGP routing algorithm (qBGP) through real-time monitoring. The simulator was created to model the emergent large scale behaviour of qBGP.

MITIE is a network topology evolution simulation tool that allows for the examination of the emergent properties of multiple co-adapting network layers. It is mostly the result of my PhD research.

Intelligent Systems
As well as the complex side of networks I am increasingly interested in Intelligent Systems, especially research topics covering knowledge analysis, semantic analysis, artificial neural networks and related techniques, such as global optimisation techniques, smart heuristics, agent-based systems and so on..