Picture of Andreas Blenk

Dr.-Ing. Andreas Blenk

Chair of Communication Networks (Prof. Kellerer)

Postal address

Arcisstr. 21
80333 München

  • Phone: +49 (89) 289 - 23540
  • Room: 0509.03.932
  • andreas.blenk(at)


Andreas Blenk studied computer science at the University of Würzburg, Germany, where he received his diploma degree in 2012. After this he joined the Chair of Communication Networks at the Technische Universität München in June 2012. In Mai 2018, he received the degree Doktor-Ingenieur (Dr.-Ing.) from the Technische Universität München with distinction (summa cum laude, Link). He is currently working as a postdoctoral-researcher and associate lecturer at the Chair of Communication Networks. His research is focused on flexible and predictable network virtualization, virtualizing software-defined networks, as well as data-driven networking algorithms.


In the following, I would like to give more details on my current research topics.

Network Virtualization (NV) & Software Defined Networking (SDN):

In software-defined networks, control plane and data plane are decoupled. Based on an open and standardized interface between both planes, network resources can be controlled, i.e., programmed, in a logically centralized manner. With networks being completely programmable, SDN enables realization of new types of network control architectures, e.g., completely distributed, semi-distributed, or completely centralized architectures. Besides, SDN makes it now possible to implement and to test these new architectures with real user traffic as it can isolate production and experimental network traffic on the same physical network. In my research, I focus on the implementation of resource management mechanisms for virtualized SDN environments --- i want to combine Network Virtualization and Software-Defined Networking to use an infrastructure as efficiently as possible given a particular use case.

Data-driven Networking & Machine Learning-based Network Design:

Whereas combining Network Virtualization & Software-Defined Networking is a step towards flexible and predictable network management, it introduces new challenges with respect to networking algorithms. Both concepts open new paths towards faster adaptations of communication networks. However, this might result in more frequent executions of networking algorithms to benefit the most from the possibilities that NV & SDN provide to manage networks. As most problems of managing networking resources are computational hard (flow routing, facility location, or the embedding of whole networks), there is a need for mechanisms that make it possible to improve algorithm efficiencies and solution qualities. In my research, i want to focus on the application of machine learning to improve networking algorithms further: e.g., networking algorithms produce valuable data of time which can be used to enhance their future executions, such as speeding up algorithms' runtimes.


In my dissertation, I focused on the virtualization of software-defined networks (available: Link).

Available from mediatum: Link

Available from Deutsche National Bibliothek: Link

Selected Publications

  • A. Blenk, A. Basta and W. Kellerer, "HyperFlex: An SDN virtualization architecture with flexible hypervisor function allocation," 2015 IFIP/IEEE International Symposium on Integrated Network Management (IM), Ottawa, ON, 2015, pp. 397-405. doi: 10.1109/INM.2015.7140316
  • A. Blenk, A. Basta, M. Reisslein and W. Kellerer, "Survey on Network Virtualization Hypervisors for Software Defined Networking," in IEEE Communications Surveys & Tutorials, vol. 18, no. 1, pp. 655-685, Firstquarter 2016. doi: 10.1109/COMST.2015.2489183
  • A. Blenk, A. Basta, J. Zerwas, M. Reisslein and W. Kellerer, "Control Plane Latency With SDN Network Hypervisors: The Cost of Virtualization," in IEEE Transactions on Network and Service Management, vol. 13, no. 3, pp. 366-380, Sept. 2016. doi: 10.1109/TNSM.2016.2587900
  • Blenk, Andreas; Kalmbach, Patrick; Schmid, Stefan; Kellerer, Wolfgang: o'zapft is: Tap Your Network Algorithm's Big Data! ACM SIGCOMM 2017 Workshop on Big Data Analytics and Machine Learning for Data Communication Networks (Big-DAMA), 2017. doi: 10.1145/3098593.3098597
  • Blenk, Andreas; Kalmbach, Patrick; Zerwas, Johannes; Jarschel, Michael; Schmid, Stefan; Kellerer, Wolfgang: NeuroViNE: A Neural Preprocessor for Your Virtual Network Embedding Algorithm. 37th IEEE Conference on Computer Communications (INFOCOM), 2018