Studentship in the Use of Distributed Data Mining for Management of Future networks

A funded PhD studentship is available for the application of distributed data mining approaches to the management of future networks . The work is, in part, conducted in conjunction with the CELTIC initiative project COMMUNE which considers cognitive management approaches to dealing with uncertainty in network management (http://www.celticplus.eu/Projects/Celtic-projects/Call8/COMMUNE/commune-default.asp) .
Communication networks play an ever more important part of modern day society and have become the basis of many diverse area of application such as health, energy, transport and even social networks. A by-product of this development has been the huge growth in the amount of data spun off by network creating at once both the need and the opportunity to distil condensed information from this great quantity of raw data through the use of machine learning and statistical techniques. Where, when and how to collect and analyse this data is a major research challenge involving trade-offs between network and computing processing costs and timeliness of information availability. The focus of the research is investigate approaches to distributed data-mining (DDM) in the context of 4G/LTE radio and broadband fiber access networks. The researcher will evaluate various algorithmic approaches in different contexts and propose architectures and criteria for placement of analysis functionality and integration of distributed result sets. The work may also consider how to deal with data uncertainty in the context of the DDM framework as well as issues around mining of data streams.