Pavlos Sermpezis, Postdoctoral Researcher, Aristotle University of Thessaloniki
Internet routing measurements: from theory to practice
Routing between networks (“inter-domain routing”) in the Internet takes place via the Border Gateway Protocol (BGP), which is a distributed, policy-based, path-vector protocol. The distributed nature of BGP, the typically unknown routing policies of networks, and the continuously changing routing decisions, make it challenging for a network to have an updated and accurate view of the routing paths that its own traffic follows. Networks are heavily based on measurements to detect events that may affect their traffic, to test whether their traffic engineering actions achieve the intended goals, etc.. In this talk, we will discuss recent advances and deployments of public measurement infrastructure that enable us to build novel Internet monitoring applications and services. Moreover, we will present our results on modeling and analyzing the Internet routing system, and how they can be used to develop theory-driven, accurate and efficient practical measurement techniques. Our experiments in the Internet reveal the gap between theory and practice: practical (naive) techniques are usually inefficient, whereas purely theory-based methods may be significantly inaccurate. We believe that bridging the gap between theory and practice can lead to efficient methodologies and evolve the Internet monitoring.
About the speaker
Pavlos Sermpezis is a postdoctoral researcher at Datalab (https://datalab.csd.auth.gr/) at the Informatics Dept., Aristotle University of Thessaloniki, Greece. Before that he was a postdoctoral researcher at the Institute of Computer Science of the Foundation for Research and Technology - Hellas (FORTH), Heraklion, Greece. He received his PhD in computer science and networks from EURECOM, Sophia-Antipolis, France (2015), and a diploma in Electrical and Computer Engineering from the Aristotle University of Thessaloniki (2011). His main research interests are in data science and computer networking with a few bits and pieces of social networks/media and recommendation systems, which he studies both from a theoretical (probability theory, graph theory, optimization) and applied (system design, measurements, experiments, data science) point of view.