Reza Rejaie "From Observations to Insight: Measuring Networked Systems under Constraints"
Speaker: Reza Rejaie
Registration: Please register via the link by 20.10.2026
Event time: 28 October 2026 14:00-15:00 (Helsinki Time Zone UTC+2) Followded by a small reception
Place: Aalto University, Maarintie 8 (TUAS) AS 2
The next lecture in the Helsinki Distinguished Lecture Series on Future Information Technology will be given by Professor Reza Rejaie. Rejaie is a Professor of Computer Science at the University of Oregon and the 2026–27 Fulbright-Nokia Distinguished Chair in Information and Communications Technologies, hosted by the University of Helsinki.
Abstract: Empirical measurement has long been an important means for gaining insight into the operation and behavior of large-scale networked systems. Yet the scale and dynamics of these systems, limitations on their visibility, and constraints on measurement resources often make obtaining representative and accurate observations challenging.
Drawing on 20 years of measurement-informed networking research, I will discuss lessons learned in two complementary settings. The first uses empirical measurement to understand foundational properties of networked systems. I will illustrate how principled sampling and inference can provide representative views from incomplete observations, and how relating observed properties to underlying mechanisms and domain knowledge helps validate their significance and extract insights that generalize beyond a particular dataset or system.
The second setting focuses on programmable network devices that enable customizable, near-real-time observation of live traffic for network management and security applications. The limited hardware resources of programmable devices must be actively managed as traffic and measurement workloads change to meet the accuracy and latency requirements of monitoring tasks. I will illustrate how monitoring systems can determine at runtime when and for how long individual measurement tasks should execute. Programmability can also provide an adaptive measurement lens, adjusting where and how measurement is focused as an event unfolds and the relevant traffic evolves. I will show how prior knowledge about structural properties of traffic can guide this search to efficiently identify traffic of interest.
These lessons collectively demonstrate that effective measurement is not defined by how much we observe, but by how purposefully we design and adapt our measurement strategy to obtain, validate, and interpret observations that yield actionable insight under practical constraints.
Bio:Reza Rejaie is a Professor of Computer Science at the University of Oregon and the 2026–27 Fulbright-Nokia Distinguished Chair in Information and Communications Technologies, hosted by the University of Helsinki. His research spans network measurement, cybersecurity, Internet architecture, peer-to-peer and multimedia networking, and social computing. He founded the Oregon Network Research Group and is a founding Associate Director of the Oregon Cybersecurity Center of Excellence. Rejaie is an IEEE Fellow (2017) and ACM Distinguished Member (2022). His honors include an NSF CAREER Award (2005) and a European Union Marie Curie Fellowship (2009). He has held visiting professor positions at Sorbonne University, Politecnico di Torino, and IMDEA Networks Institute. Before joining the University of Oregon, he was a Senior Technical Staff Member at AT&T Labs–Research. He received his M.S. (1996) and Ph.D. (1999) in Computer Science from the University of Southern California (USC).