Digital signal processing is the technological driver for energy-efficient and high-performance optical transport networks. It increases the robustness against transmission interferences and enables software control of the physical layer.
Electronic Digital Signal Processing (DSP) is a key technology for optical transport networks, in particular for coherent optical transmission systems. In optical transponders, it enables carrier recovery and synchronization as well as compensation of linear and non-linear signal interference. In addition, various signal and distance parameters can be estimated and made available as valuable telemetry data of the control layer in optical transport networks. Methods of machine learning are increasingly being used for these applications.
The working group Digital Signal Processing has extensive experience in the development and real-time implementation of DSP-algorithms for optical communication systems. Our competencies range from the design of algorithms to their implementation in different programming languages such as Matlab, Python, and C/C++ to VHDL-based FPGA implementations and are regularly proven in laboratory experiments and field tests. A particular focus is on the modeling of nonlinear systems, corresponding estimation algorithms and the design of nonlinear equalizers. Our research activities in the field of digital signal processing for optical communication systems focus on the following aspects:
Development of algorithms for:
- Carrier recovery and synchronization for advanced modulation techniques
- Linear and nonlinear channel estimation
- Compensation of linear and non-linear signal interference
- Linearization of electrical and optical components
- DSP-based performance monitoring and component characterization
- Machine learning