TU Berlin

Fachgebiet Theoretische Grundlagen der Kommunikationstechnik15 January 2016 - Dipl.-Ing. Martin Kasparick (Fraunhofer Heinrich Hertz Institute)

Schriftzug CommIT


zur Navigation

Invitation To A Talk by Dipl.-Ing. Martin Kasparick, Fraunhofer Heinrich Hertz Institute (HHI), TU Berlin

15 January 2016, 11:00 AM
HFT - Hochfrequenztechnik building, 6th floor, 
Room HFT-TA 617, Einsteinufer 25, 10587 Berlin
Adaptive Learning For Self-Organizing Communication Networks


Due to their dense and heterogeneous nature, the scarcity of resources, and high costs of manual configurations, the ability to self-organize is considered to be an essential feature of the next generation of wireless networks. Most self-organization mechanisms need a certain degree of knowledge on the system state, but explicit measurements are often infeasible. This talk deals with adaptive learning techniques that enhance wireless networks with the ability to extract the required information from user measurements and available side information. In particular, we address the problem of reconstructing radio maps in an online fashion. We propose and evaluate kernel-based algorithms for online learning that overcome many shortcomings of frequently used batch learning methods. The complexity is significantly reduced by imposing sparsity to limit the amount of data that is saved and processed. Based on different data sets, we evaluate the performance of the methods in applications, such as path-loss and interference estimation.


Martin Kasparick received the Dipl.-Ing. degree in computer engineering from the Technische Universität (TU) Berlin, Germany, in 2009. From 2010 to 2013 he has been an research associate at the Heinrich-Hertz Chair for Information Theory and Theoretical Information Technology at the TU Berlin and since 2013 he is an research associate with the Fraunhofer Institute for Telecommunications, Heinrich Hertz Institute (HHI) in Berlin. Currently he is finalizing his Ph.D. at the TU-Berlin on self-organizing networks. His research interests are in the area of adaptive learning and optimization in wireless networks.



Schnellnavigation zur Seite über Nummerneingabe