http://hdl.handle.net/10900/124616 Tue, 12 May 2026 19:27:46 GMT 2026-05-12T19:27:46Z https://publikationen.uni-tuebingen.de:443/xmlui/bitstream/id/572ebe39-9ac1-4c08-ab6a-7cb6182e56df/ http://hdl.handle.net/10900/124616 http://hdl.handle.net/10900/126092 Project Overview: Router for Academia Research & Education (RARE) Loui, Fréderic; Mate, Csaba; Gall, Alexander; Wisslé, Maxime In this paper, we give a brief overview of the GÉAN project Router for Academia Research & Education (RARE). Its objective is to create a modern open routing platform. It leverages the FreeRtr control plane and programmable data planes that aims to support Research and Education (R&E) specific use cases. RARE/freeRtr provides different data plane implementations that allow for a high degree of deployment flexibility. Most router features are developed with P4 for the software target BMv2, and where possible also for the Intel Tofino hardware target. Additionally, a DPDK-based data plane called P4DPDK for x86 platforms is also available that addresses use cases requiring lower bandwidth. Thu, 07 Apr 2022 00:00:00 GMT http://hdl.handle.net/10900/126092 2022-04-07T00:00:00Z http://hdl.handle.net/10900/126091 Are Kubernetes CNI solutions ready for >10 Gbit/s? Volpert, Simon; Eisenhart, Georg; Domaschka, Jörg The growth in variety of Container Network Interface (CNI) implementations as well as increasing Kubernetes adoption on bare metal require a thorough assessment of its performance characteristics. Specifically, the question whether CNI can saturate network links above 10 Gbit/s is of interest. This paper highlights an initial approach and central research questions towards the goal of this assessment. Therefore, we propose a method to acquire measurements that act as a groundwork for further investigations. Preliminary results show that a simple setup is not able to saturate even a 50 Gbit/s link. Thu, 07 Apr 2022 00:00:00 GMT http://hdl.handle.net/10900/126091 2022-04-07T00:00:00Z http://hdl.handle.net/10900/126090 Recognition of Similar NetFlow Data in Decentralised Monitoring Environments Eisenhart, Georg; Volpert, Simon; Braitinger, Jan; Domaschka, Jörg One of the main challenges in the analysis of NetFlow data in decentralised monitoring environments comes from merging datasets from different independent sites. One problem is to identify similar data points which can impact derived metrics from such data directly. This article provides a proof of concept how similarity measurements based on distance metrics can be used to identify similar or related flows from different datasets. For this, several domains are outlined which can benefit from this approach to support validation of research scenarios and data analysis. Thu, 07 Apr 2022 00:00:00 GMT http://hdl.handle.net/10900/126090 2022-04-07T00:00:00Z http://hdl.handle.net/10900/126089 Procedure-based Functional Decomposition for 5G Core Network Functions Goshi, Endri; Stahl, Raffael; He, Mu; Pries, Rastin; Kellerer, Wolfgang The adoption of the cloud-native paradigm in mobile core networks has introduced increased traffic in the controlplane stemming from the inter-Network Function (NF) communication. In this paper, we evaluate a different functional (de)composition for the 5G Core NFs with the goal of achieving high performance and efficient orchestration while reducing the inter-NF traffic. Thu, 07 Apr 2022 00:00:00 GMT http://hdl.handle.net/10900/126089 2022-04-07T00:00:00Z