• de
Prof. Dr. Holger Karl

Softwarized deployment of services in waves around moving users (SWAVES)

Modern software is often designed and distributed as a collection of cooperating components - examples are microservice chains or network function virtualization. this can bring significant advantages in edge computing scenarios when components can be run close to a user. But once users start moving around, we need to make sure that the right component is available and running at the right location - we need to solve a mobile lifecycle management challenge.

The DFG-funded project SWAVES addresses this challenge. We will research how to distribute exectuable artefacts (e.g., source code, containers, virtual machine images) to edge computing sites, to transition them between different execution states (e.g., not running, cold or warm standby) and how to move state around between them. This requires complex decisions and should benefit greatly from mobility predictions. To address these issues, we will look at both conventional techniques as well as at machine learning-based approaches.

The research project will be undertaken in collaboration with Mercator fellows Carla Fabiana Chiasserini and Claudio Errore Casetti, both from Politecnico di Torino. 

Job offer: Open position for a PhD student (fully-funded E13-style position) plus student assistant. Email Holger Karl if interested. 

Investigators: OPEN POSITION!
Project duration: January 2023 - December 2025
Funded by: Deutsche Forschungsgemeinschaft (DFG) 

open6GHub: 6G für Menschen, Umwelt & Gesellschaft

open6GHub: 6G for people, environment, & society

open6GHub is a large research project, funded by the German federal government via the Ministry of Education and Research; it is coordinate by DFKI and comprises 14 partners all over Germany. We contribute mostly to two aspects: 

  • Networked Intelligence: How and where can machine learning-based artifical intelligence techniques be used to improve network operation, as well as how can networks assist in running such ML techniques in a distributed fashion, for both learning and inference phases? 
  • Programmable infrastructure: How can an infrastructure be managed when the infrastructure itself becomes adaptive; e.g., how to orchestrate distributed execution of microservice chains inside an edge cloud when the edge comprises mobile base stations (e.g., drones) with local computing capacity. Also, how can dependability be improved when operating a software-defined network. 

This project has several open positions for student research assistants, working on topics like a testing/emulation infrastructure for netowrk softwarization, dynamic infrastructure, or ML in networks. Do contact us if interested! 

Investigators: Valentin Kirchner, Leonard Paeleke 
Project duration: July 2021 - June 2025 
Funded by: Bundesministerium für Bildung und Forschung (BMBF) 
Publications: Separate page 

Acoustic Sensor Networks

The idea of this project is to use distributed microphones, connected together by wireless communication, for sophisticated acoustic applications. Examples include detecting the location of a speaker, echo cancellation, denoising a recording, or selecting the best microphone for alternating speakers. These acoustic applications can be useful in use cases like class rooms, conference or lecture halls, or meeting rooms. 

A challenge here is that these applications require high-resolution audio streams, resulting in high data rate, possibly over longer distances (e.g., in conference halls). In addition, latency should be low. To support these requirements, it is promising to distribute signal-processing functions directly onto the microphones (or possibly, intermediate nodes). This can reduce data rates and improve latency and overall quality of experience. 

This idea is investigated as project P1 of the DFG Research Group (DFG Forschergruppe) FOR 2457 Acoustic Sensor Networks


Investigators: Haitham Afifi
Project duration: February 2017 - January 2023 (extended after three years) 
Funded by: Deutsche Forschungsgemeinschaft (DFG) 
Publications: Separate page

CRC 901: On the fly computing

This is a research project funded by the DFG collaborative research center SFB 901 "On-The-Fly-Computing". The research center investigates individually and automatically configured IT services which are composed from and executed on a marked of world wide traded combinalbe services and execution platforms. Within this research center our research group investigates the the communication network as one important ingredient for such market platform. Our research covers the realization and optimization of overlays over real networks as well as the optimized allocation of resources in single- and multi-site data centers. More details can be found on adedicated web page

Investigators: Adrian Redder (and in the past: Asif Hasnain, Arne Schwabe, Philip Wette)
Project duration: July 2011 - June 2023 (with extensions after four and eight years)
Funded by: Deutsche Forschungsgemeinschaft (DFG) 
Publications: Separate page