PhD assistants (wissenschaftliche Mitarbeiter)
There are currently openings in several projects and topics.
- Project SWAVES: How to deploy executable software artefacts in mobile edge computing? Using both conventional and machine-learning based approaches, decide which container/VM image/... needs to be pushed where, in which form, to make sure that services can run without interrupts when users move around in a network. [Official job offer (in German)]
- Project KI-Servicezentrum Berlin-Brandenburg/Service centre for artificial intelligence: How to manage resources in machine-learning environments, in particular, when both infrastructure and workloads are heterogeneous (e.g., combining batch and interactive ML jobs for both training and inference or non-ML jobs).
- Project Data Analytics for Cloud Operation: How to do predictive maintenance for cloud systems, based on machine-learning models trained from cloud operation logs? [Official job offer]
- Project Open6GHub: What happens when reconfiugrable networks meet microservices on network function virtualization? How to align degrees of freedom from network reconfiguration with modifying the software service it is executing?
Open positions funded by the group: Independent of a specific external project, these positions are usually flexible with respect to the research topic but do carry a teaching obligation. [ PDF (de) ][ PDF (en) ]
Possible topics are (more details on the group description page):
- Network softwarization and microservices: Networks (like 5 and 6G as well as the Internet at large) are developing towards a general-purpose, programmable infrastructure. Questions in this field can range from software architecture and description techniques over algorithmic aspects (optimization, heuristics, machine learning) for many variants of management/control problems, to business models and system design.
- 6G research in particular: For 6G networks, we will look at programmable infrastructures and the use of machine learning to optimize such networks, as well as the deployment of machine learning inside a 6G network. We seek at least two PhD students her!
- Mobile networks: While related to wireless control, mobile networks take a broader perspective, e.g., optimizing resources during handover. This is an area where machine learning has delivered interesting first results, but the step to real applicability (eg., true network-wide scalability) is still missing.
- Data centre networks: Data centres carry very characterisitc work loads, which should be properly managed. In addition, they also have become programmable, e.g., by integrating programmable NICs with server hardware. In the end, typical data centre workloads like map/reduce should become more efficient.
These positions are intended for students during their Bachelor or Master program, in parallel to their ongoing studies.
Student research assistants
If you are interested in joining network-related research projects, recently often using machine-learning techniques, please do talk to us. Possible topics include, but are not limited to:
Bring Chaos into Networks: Today's networks are increasingly programmable, requiring software engineering practices. Software engineers build, test and deliver software automatically (CI/CD) and test software under turbulent conditions (chaos engineering). Your job is to integrate these practices into an light-weight, open-source testbed, that is based on Linux processes, Docker containers, and Python.
If you are interested in software engineering practices, like to code in Python, feel familiar with the shell and have interests in basics of Linux processes and Linux networking, then this job will be fun.
- Laying ground work for the upcoming Wireless Control Lab at HPI: would it be interesting to control a swarm of Crazyflies wirelessly?
- Wireless protocols custom-tailored to control applications: 802.11 does not cut it? Let's build a testbed to experiment with our own variants
- Systematic profiling of network functions or microservices, including the build process: Contribute to the existing SNDZoo
- Distributed machine learning: does quality matter when resources are not invite?
We are open to wide range of interesting topics - drop by for a chat (in person or online): send email.
Student teaching assistants
Depending on the teaching plans for the current or upcoming semester, there might also be opportunities for teaching assistants (Tutoren). Please see the teaching page for more information.