Alexander Dietmüller

I am a networking researcher focused on combining machine learning and control theory with communication networks. For me, this means: (i) how can we learn the interactions between traffic and networks; and (ii) how can ensure that the data-driven decisions are optimal, in particular at the tail?

I defended my dissertation titled Adaptive Network Traffic Modeling on the 29th of July 2024 under supervision of Prof. Dr. Laurent Vanbever, co-examined by Prof. Dr. Arpit Gupta and Dr. Francis Yan.

Among other topics, my dissertation tackles the problem of keeping ML models up to date with a focus on tail performance. (Re-training) ML models over time is known as continual learning (CL), and common CL systems are mostly focused on delivering good performance on average. Yet in networking, tail performance is very important, and thus I developed Memento: a system to reliably identify and remember rare samples to improve model performance at the tail. Furthermore, I also investigated how to efficiently learn general network dynamics by leveraging state-of-the-art ML architectures like Transformers to generalize network dynamics, i.e. extracting general patterns from a variety of traces for future predictions.

In 2022, I worked as a research intern with the Measurement Lab team at Google, supervised by Prof. Dr. Phillipa Gill and Matt Mathis.

In the past, I worked on programmable packet scheduling together with Dr. Albert Gran Alcoz and on using programmable data planes to learn traffic feature distribuitions together with Dr. Georgia Fragkouli. Finally, and with the help of other Master students, I also investigated various aspects of congestion control.

I received my Bachelor and Master degrees in Electrical Engineering and Information Technology from ETH Zürich.

Talks

Publications

Supervised Theses