Unpredictable, large-scale and complex. Familiar words to members of the NETWORKS consortium; these are the characteristics of the networks that they study and improve.
UvA professor of Applied Probability, Michel Mandjes, is project leader of NETWORKS and describes its objective. 'We want to use networks with scarce resources as effectively as possible, taking uncertainties into account. The use, for example, of a communications network depends on the unpredictable behaviour of its users and other coincidental elements such as failures. We cope with this uncertainty by incorporating these factors into our models.'
Besides in telecommunications, NETWORKS aims to build self-organising, intelligent networks for transport, energy networks and road traffic. These are all areas chosen for their social importance, says Mandjes. 'NETWORKS is fundamental in nature but its underlying objective is to actually apply the knowledge acquired.’
NETWORKS will run for ten years, the first of which has already expired. During this year, Mandjes and his colleagues established the structure and defined the themes for PhD research. They also reflected on how they could really bring the participants together. Mandjes: 'People tend to stick to their own area, but that's not our intention. That's why we organised it so that many PhD projects come under two supervisors who have not worked together before.’
He will personally supervise four or five PhD students as part of NETWORKS. Two of these are funded by the VU. UvA is one of the four core members of NETWORKS; VU is an affiliate member and contributes accordingly. Mandjes is thrilled about this additional collaboration with his VU colleagues. Although researchers at both universities partly study the same themes, there is very little overlap, he explains. 'This is an opportunity to learn about areas that border on your own expertise.
Mandjes and his VU colleagues are recruiting for one more PhD project. ‘It focuses on managing congestion in road traffic networks. This obviously conjures up matters such as optimal positioning and operations for traffic lights but there are also more sophisticated techniques that regulate inflows, as well as matrix signs with speed recommendations.' Mandjes considers the composition of the supervisory committee to be very interesting. 'There is a mathematician/physicist on board that approaches the problems with genuine models of statistical mechanics. Then there is someone that addresses uncertainty using operations research and I predominantly take the perspective of applied probability.’
He believes it is remarkable that they were fortunate enough to obtain funding for a consortium involved in mathematics, which is often viewed as a 'serving' science. 'We all think science is amazing, wonderful and powerful. The fact that we so appreciate the elegance and abstraction of mathematics is what binds us together.'