Research to get things moving!
Faculty Prize goes to Signal Processing PhD Researcher
We are delighted to announce that one of our PhD students has been awarded the annual Graduate School Research Student Prize for the Engineering Faculty. The prize of £400 is awarded for outstanding academic performance and academic achievement, an accolade in our research-rich engineering community of nearly 800 research students.
Amod Jai Ganesh Anandkumar, working in our Advanced Signal Processing Group, will be completing his thesis this year and is busy writing up but took time to explain what his research is investigating.
“There has been a huge demand for frequency space allocated for the use of wireless networks in recent years. This is sure to continue rising due to the great increase in broadband internet access using portable devices such as laptops, netbooks, tablets and smartphones. Currently, these wireless networks have to make do with an allocation of only a small fraction of the usable frequency spectrum, competing with various other military and commerical applications. However, studies have shown that this seeming shortage of spectrum is not really the case in practice. Even Manhattan, one of the busiest and most crowded places in the world in terms of frequency spectrum use, has been observed to achieve an efficiency of only about 13% and this is primarily due to heavy cellphone and wireless network usage.”
Lobbying for a more generous government allocation of frequency space is one solution; another answer lies in the focus of Amod’s research.
“A common feature of many prospective paradigms for next generation wireless networks has been the concept of dynamic spectrum access. The idea is to use licensed spectrum allocated to other applications in a way that does not interfere with their operation. This would allow these next generation networks to achieve additional capacity without increasing the requirement for costly frequency spectrum. In such systems, we have to design algorithms to optimally use all these spectra of varying quality to achieve the highest capacity with minimal resources.”
Amod uses a road-traffic analogy to explain: “If you imagine a multi-lane highway where each lane has a different speed due to different levels of traffic and you want to make deliveries with a fleet of trucks. It is possible to work out a way to spread the trucks across the lanes in order to achieve the best overall delivery time. Simply using the fastest lane for all trucks will not work out to be the best, as the additional traffic of your own trucks may end up slowing the lane. The best answer will be a specific way of spreading your traffic across all the lanes to achieve the best possible solution.”
In a similar way, data can be routed across channels of differing bandwidth to obtain maximum efficiency. Amod is working out this optimisation of the allocation of available resources using a field of applied mathematics called game theorywhich aims to capture, analyse and predict the behaviour of players in complex strategic scenarios.
“To continue the analogy, there may be many transport companies using the same highway for their own fleet of trucks. This leads to a competitive scenario with each company operating independently to maximize their own profits. Game theory provides many tools to design and characterize algorithms that can solve such multi-user problems. A major consideration in such cases is that the system remains stable and available (a highway with no collisions, crashes, lane closures or traffic queues caused by rapid and non-stop lane changes!)”.
“One of the key aspects of my research was to design these algorithms so that they could successfully function even with imperfect information being fed into the system. This was important because errors are always present in practical measurements and there had been no substantial investigation into the effects of such errors and ways to mitigate them.”
This is a global challenge and Amod has established his research group at Loughborough University as an internationally-recognised centre of excellence in his research area through working with Professor Bjorn Ottersten and his team at KTH Royal Institute of Technology, Stockholm, Sweden; Dr. I. Menache, Microsoft Research, California, USA and Dr. A. Anandkumar, University of California Irvine, USA.
Amod’s supervisors, Professor Jonathon Chambers and Dr Sangarapillai Lambotharan, rate Amod very highly and remark “Amod is a truly outstanding PhD student. Having delivered world-leading research findings in the area of resource allocation for multi-user wireless communications through the exploitation of Game Theory and robust optimization and presented papers in the foremost conferences in his field, he truly merits this prize.”
 |
Amod receives his award from Will Spinks, the University's Chief Operating Officer.
|