Tuesday 25.08, 13:30 - 14:30
QoE: from Measurements to Business
Dr. Martin Varela, VTT Technical Research Centre of Finland.
Quality of Experience (QoE) has become a trending topic in networking, multimedia and other research domains. In this talk ew will discuss QoE's role as the ultimate performance metric from networks: how do users perceive network service performance?. With this in mind, we will go through basic concepts related to QoE estimation and how it can be exploited from the technical and business perspectives, as well as its requirements in terms of network metrology..
About the Speaker::
Martin Varela received his PhD and MSc from the University of Rennes 1 (Rennes, France), in 2005 and 2002 respectively. He has been an ERCIM fellow, and spent time at SICS and VTT, where he is currently a Principal Scientist. He has recently been a guest Senior Researcher at the Cooperative Systems Group at the University of Vienna. He leads VTT's work on QoE, and was a Finnish management committee member for the recently ended COST Action IC1003 Qualinet. He is currently co-chair for the IEEE MMTC QoE Interest Group. His research interests lie in the QoE domain, in which he has been active since 2002, with a particular focus on real-time QoE models for generic services, and applications thereof. He is particularly interested on issues related to QoE-driven cross-layer control, resource management and business applications.
Wednesday 26.08, 10:30 - 11:30
Network traffic classification: From theory to practice
Prof. Pere Barlet-Ros, UPC BarcelonaTech, Spain.
Network traffic classification is an important problem for network operation and management but challenging from a research perspective. Although the research area of traffic classification has been very prolific during the last decade, in this talk we will argue that this problem is still far to be completely solved. For example, existing research proposals are difficult to validate and compare, while they present several practical limitations that hinder their actual deployment in production networks. In this talk, we will review the traffic classification problem from a more practical perspective. We will discuss the limitations of previous research proposals and present an autonomic traffic classification system that addresses them. We will also describe our experience in turning our research prototype into a commercial-grade product (Network Polygraph), which already counts with several customers worldwide (see: https://polygraph.io).
About the Speaker::
Pere Barlet-Ros (http://people.ac.upc.edu/pbarlet) received the M.Sc. and Ph.D. degrees in Computer Science from UPC BarcelonaTech in 2003 and 2008 respectively. He is an Associate Professor with the Computer Architecture Department at UPC BarcelonaTech, and the co-founder and Chairman of Talaia Networks (http://www.talaianetworks.com). He was also a visiting researcher with Endace, New Zealand (Winter 2004), Intel Research Cambridge, UK (Summer 2004) and Intel Labs Berkeley, California (Summer 2007). He has recently co-chaired TMA 2015 and served in the TPC of several conferences and workshops in the area of network measurements. His research interests are in the fields of network monitoring, network data analysis, traffic classification, anomaly detection, SDN measurements and online privacy.