Research: Chair of Communications Engineering

Coding and Modulation

Power efficient communication requires higher order modulation and forward error correction (FEC) codes. We are working on pushing communication to the theoretical limits by implementing Shannon‘s blueprint. Two key concepts are distribution matching and probabilistic shaping, which combine non-uniform signaling with FEC. We design and implement state-of-the-art low-density parity-check (LDPC) and polar codes over binary and non-binary fields. We test our prototypes in close collaboration with industry.

Currently working in this area:

  • Emna Ben Yacoub
  • Mustafa Cemil Coşkun
  • Delcho Donev
  • Thomas Jerkovits
  • Marcin Pikus
  • Tobias Prinz
  • Patrick Schulte
  • Fabian Steiner
  • Thomas Wiegart
  • Peihong Yuan

Compressed Sensing, Machine Learning, Security

We are developing information theoretic frameworks, codes, and algorithms for signal processing problems such as machine learning and compressed sensing. As the need for secure communications increases, we concentrate on privacy and secrecy related topics such as physical unclonable functions

Currently working in this area:

  • Ali Rana Amjad
  • Onur Günlü
  • Lars Palzer

MIMO and Massive MIMO

The demand for higher data rates via wireless channels continues to grow. Base stations and terminals with many antennas are essential to meet the demand, and massive MIMO considers hundreds or thousands of antennas with simplified signal processing. Our research focuses on practical implications such as EIRP constraints and complexity reduction via few-bit analog-to-digital converters. We further try to consider physically correct modeling and realistic device constraints.

Currently working in this area:

  • Amir Ahmadian
  • Andrei Nedelcu
  • Markus Staudacher

Optical Communications

Fiber optic cables form the backbone of our global communication networks, no other media allows a faster data exchange. A useful description of waveform propagation in optical fiber is given by the non-linear Schrödinger equation (NLSE). However, its form does not admit easy insight of information theoretically relevant quantities such as capacity. Recently, the Non-Linear Fourier  Transform (NFT) was suggested as a means to explain some of the effects observed before. We are investigating this approach.

Currently working in this area:

  • Javier Garcia