A Forum Numerica Webinar
by Erol Gelenbe
Professor of the Institute of Theoretical and Applied Informatics, Polish Academy of Sciences
Honorary Professor University of Electronic Science and Technology of China (UESTC)
& Associate Researcher I3S, Université Côte d'Azur

Abstract:
We will review our work on the design of packet networks which adapt their routes using reinforcement learning, based on on-line measurements and random neural networks that are distributed at selected routers. The choice of network paths is driven by an objective or "goal" which includes broad quality of service criteria such as delay, loss, energy and security. The implementation remains compatible with the IP protocol. Different experimental results will be presented on several test-beds, as well as in the framework of on an intercontinental overlay network, and also separately through an implementation in SDN. We will also outline how similar ideas have been tested for task allocation in the Fog or Cloud. The work has resulted in several PhD dissertations and post-doctoral projects, the FP7 and H2020 SerIoT project, a few industry patents, and publications in mainstream journals and conferences (Proceedings of the IEEE, Communications ACM, IEEE JSAC, ICC, MASCOTS, IEEE Trans. Cloud Computing, etc.).

Speaker's Bio:
Born in Istanbul and a graduate of METU (Ankara), Erol Gelenbe founded the performance analysis and modelling research at INRIA and in France's universities. He is a Fellow of both IEEE and ACM, and was a Professor in France at Université Paris-Nord, Université Paris-Sud Orsay, Université René Descartes (Paris V) and Ecole Polytechnique. He co-founded the DEA (graduate program) and the first Computer Science CNRS Laboratory "Al Khowarizmi" at Université Paris-Sud (Orsay). Known for inventing the Random Neural Network, its Learning Algorithms, stochastic networks known as G-Networks which have product form solution, and Energy Packet Networks, he has also contributed in the last decade to the theory of gene regulatory networks and quantum communications. He has graduated over 95 PhDs in France, Belgium, the USA, and UK, and served as Chaired Professor at the Université de Liège (Belgium), Department Head of Electrical and Computer Engineering at Duke University, Director of the School of EECS at the University of Central Florida, and Head of Intelligent Systems and Networks at Imperial College London. He was elected Fellow IEEE, Fellow ACM, Fellow of the Académie des Technologies (France, 2008), the Royal Academy of Sciences, Arts and Letters of Belgium (2015), and the Science Academies of Hungary (2010), Poland (2013) and Turkey (2007). After Brexit he moved his EU H2020 projects to the Polish Academy of Sciences, retiring from Imperial College in early 2019 and becoming a full-time Professor of the Polish Academy of Sciences. He was awarded Chevalier de la Légion d'Honneur (2014) and Commandeur du Mérite (2019) by France. Italy awarded him Commendatore al Merito della Repubblica (2005) and Grande Ufficiale dell'Ordine della Stella d'Italia (2007). He won the Parlar Foundation Science Award (Ankara, 1994), the Grand Prix France Télécom (now Orange) of the French Science Academy (1996), the ACM SIGMETRICS Life-Time Achievement Award (2008), the Oliver Lodge Medal (IET, UK, 2010), the Dennis Gabor Award of the Hungarian Academy of Sciences (2013), and Honoris Causa doctorates from University of Rome (Italy, 1996), Bogazici University (Istanbul, 2004), and University of Liège (2006) . He currently works on energy efficiency and security together with quality of service in systems and networks using machine learning and mathematical models, as part of three ongoing H2020 projects: SerIoT (as coordinator), SDK4ED and IoTAC. He is also Scientific Advisor to the Huawei Dublin Research Center. He was recently elected by the IFIP General Assembly as one of its first 18 Fellows worldwide, as part of IFIP's 60th Anniversary Celebrations in 2020.