Close this search box.

Best Paper Award for Cyprus-Based Research Team


An innovative framework for the control and the management of telecommunication networks

The Best Paper Award was given at a major international scientific conference to a Cyprus-based team consisting of researchers from the KIOS Research and Innovation Center of Excellence and the Department of Electrical and Computer Engineering at the University of Cyprus (Research Fellow Dr Tania Panayiotou and Associate Professor  Georgios Ellinas) and the Department of Electrical Engineering, Computer Engineering, and Informatics at the Cyprus University of Technology (Assistant Professor Sotirios Chatzis). The awarded paper presented a research effort that aims to improve the performance of optical networks which could potentially reduce the operational and capital expenditure for telecommunication network providers.

This award demonstrates the excellent quality of research which is being undertaken in Cyprus and is a significant international recognition for the research performed in the field of Information and Communications Technologies (ICT) and in particular in the area of Critical Infrastructure Systems.

The award winning paper entitled “A Probabilistic Approach for Failure Localization” was presented at the 21th IEEE International Conference on Optical Network Design and Modeling (ONDM) in Budapest, Hungary, in May 2017.  The IEEE International Conference addresses cutting-edge research in established areas attracting significant interest from several international leading research centers from universities and industry. The scientific conference explores novel and emerging topics in the areas of optical networking, optical systems and optical network architectures, wireless optical networks, photonic integrated networks, as well as control and management developments in optical networks.

The award-winning paper addresses the critical problem of fault localization in transparent optical networks. In such networks localizing network failures (i.e., fiber cuts, equipment failures, etc.) is not trivial. Usually, upon a failure incident, network technicians are called for real-time localization of the failure using monitoring data. The mean time to repair (MTTR) the failures can vary from several hours to days depending on where the failure actually occurred. Reducing the MTTR is of critical importance as network failures may severely affect the network availability and can cause severe service disruption.

The novelty of the award-winning paper lies in an innovative framework for localizing network failures in an automated fashion, with the use of advanced statistical machine learning techniques. The proposed approach can be trained on historical datasets that describe past failure incidents for fast and accurate localization of current failure events.  This approach minimizes human intervention and/or the use of specific monitoring equipment. This novel framework can be used by network operators to reduce the MTTR, the human effort required, and the number of monitoring equipment. Therefore, network providers could reduce their Capital and Operational Expenditures (CAPEX/OPEX) related to the fault localization procedure.