We look forward to receiving your emails to supervise you on “hot topics” of communications engineering, which should enable you to argue in the future that you have a solid background in this area to prepare you for your future career – both in academia and in industry.

If you are interested in any of these (or similar) topics that combine communications engineering with machine learning and cybersecurity, you can send an email to Prof. Dr.-Ing. Onur Günlü via onur.guenlue@tu-dortmund.de with your background and academic interests (possibly with your CV and transcript). If you are not yet sure, you can just write an email and we will help and guide you clearly to help you succeed!

We have various thesis and project topics for Bachelor level, some of which are listed below, and we are open for suggestions and discussions! In general, communications and computations over wireless channels are our main directions and the details are to be adjusted according to the strengths of the student. A typical thesis and project consist of reading assigned papers, summarizing the state-of-the-art based on the assigned papers and others, and simulating (in software) or implementing (in hardware) basic use cases of the topic. Preliminaries for each topic vary a bit, but in general attending the Nachrichtentechnik I or a similar course, combined with knowledge in either Python or Matlab, will suffice.
 

• Secret Key Generation from Wireless Channels for 6G
  The wireless channel has random noise in it, mainly due to the receiver hardware impairments. This random noise can be used to extract secret keys shared between a legitimate transmitter and receiver by using 6G communication systems, which is the main high-level direction for this topic.
   
• Computing Functions via Wireless Communications
  The wireless channel, especially when the noise levels are low, can be used to “do matrix multiplications directly over the air”, which can bring immense energy benefits, which will be considered in this project.
   
• Distributed Machine Learning Methods under Privacy Constraints
  Machine learning methods require high computing power, which is not always available. Therefore, the computations can be distributed among multiple computation nodes, which triggers privacy concerns. We will consider privacy-aware distributed machine learning methods to make the communication between distributed computation nodes private.
   
• Compressing Images with Neural Networks for Realistic Image Generation
  Realistic image transmission does not require losslessly reconstructing each pixel value of an image, which enables communication engineers to send extremely low-energy data for image reconstruction at another node. This project considers both semantic image compression (i.e., only send the most important semantic information) and generative artificial intelligence methods for image reconstruction, which provide immense gains in communications.

Possible Topics for Master Thesis & Master Projects at TUDO NT

We have various thesis and project topics for Master level, some of which are listed below, and we are open for suggestions and discussions! In general, communications and computations over wireless channels are our main directions and the details are to be adjusted according to the strengths of the student. A typical thesis and project consist of reading assigned papers, summarizing the state-of-the-art based on the assigned papers and others, proposing fundamental directions to advance the state-of-the-art together with your supervisor, and simulating (in software) or implementing (in hardware) relevant use cases of the topic to illustrate fundamental performance gains. Preliminaries for each topic vary a bit, but in general attending the Nachrichtentechnik I or a similar course as well as Digitale Quellenkodierung and/or Secure Communications courses, combined with knowledge in either Python or Matlab, will suffice.

• Deep Learning for Code Constructions for 6G
  Deep neural networks will be used as error-correcting codes in 6G systems.
   
• Integrated Sensing and Communications for Future Networks
  Using the same signal for communicating with a receiver and for sensing nearbytargets is a highly promising topic that will be implemented in 6G.
   
• Private Distributed Computing for AI Datacenters
  Distributed matrix multiplication is a fundamental requirement for distributed machine learning and other important tasks, which requires privacy-aware communication schemes.
   
• Secure Communications with Shannon Theoretic Methods
  Provable security can be established by using fundamental Shannon-theoretic methods and we work on numerous applications of them.
   
• Secure Communications with Post-Quantum Cryptography
  Quantum computers will make our current encryption systems vulnerable, so we focus on post-quantum cryptographic methods that provide security against computationally powerful attackers.
   
• Neural Compression Methods
  Using neural networks for compression of data has many new applications, which will be considered in this project.