Welcome to UPB-CN!
We conduct research and teaching in networked systems of all kinds, ranging from cloud data centers to mobile devices.
With the rapid growth of AI workloads, networked systems have become critical infrastructure for training and serving large machine learning models. Our research explores how to make such systems more scalable, efficient, and sustainable.
The computer networks group is part of the Department of Computer Science at Paderborn University.


Latest News
All news
Paper Accepted to DATE 2027
Our paper titled “Optimizing LLM Inference on Heterogeneous Edge SoCs with Operator-Level Controls” has been accepted to the 30th Design, Automation and Test in Europe Conference (DATE 2027), to be held in Dresden, Germany during March 22-24, 2027.
Paper Accepted to ACM SoCC 2026
Our paper titled “FluxReduce: Straggler-Resilient In-Network Aggregation for Distributed Machine Learning” has been accepted to the 17th Annual ACM Symposium on Cloud Computing (SoCC 2026), to be held in Singapore during November 18-20, 2026.
PhD Defense of Marvin Illian
Marvin Illian successfully defended his PhD thesis titled “Adaptive Connectivity Management in Heterogeneous Mobile Networks”.
IEEE WiOpt 2026 Best Paper Award
Our paper titled “Cells on Autopilot: Adaptive Cell (Re)Selection via Reinforcement Learning” won the Best Paper Award at the 24th International Symposium on Modeling and Optimization in Mobile, Ad hoc, and Wireless Networks (WiOpt) held in Columbus, OH, USA.Selected Publications
All publications- DATE 2027 Optimizing LLM Inference on Heterogeneous Edge SoCs with Operator-Level Controls
- SoCC 2026 FluxReduce: Straggler-Resilient In-Network Aggregation for Distributed Machine Learning
- EuroMLSys 2026 Before the First Token: Benchmarking Data Preprocessing in Vision-Language Models
- MLSys 2026 Breaking the Ice: Analyzing Cold Start Latency in vLLM
- MobiCom 2025 Uirapuru: Timely Video Analytics for High-Resolution Steerable Cameras on Edge Devices
- SC 2024 NetCL: A Unified Programming Framework for In-Network Computing