Topics for theses

Contact: application (at) Cyberagentur.de

High-quality data forms the basis for efficient machine learning algorithms. However, such data is not always available in the required quantity. In addition, obtaining the relevant data can be a very time-consuming and resource-intensive process. In order to develop flexible and resource-efficient machine learning models, there are learning methods that make it possible to achieve promising results with just a few data examples (e.g. images, videos, text modules). Such learning methods are also known as low-shot methods and are characterized by a data-saving training process.

The aim of a master’s thesis is to investigate the extent to which low-shot learning methods (few-shot, zero-shot and/or in-context learning) can be used for classification tasks in safety-relevant contexts. To this end, the current state of research in this area will be examined and described, various low-shot classification techniques will be compared and a prototype will be developed that can be used as a demonstrator for classification tasks.

The basis of the applied part of the master’s thesis is a data set that includes image data of different classes of military vehicles and can be used for an exemplary implementation of the methods developed.

Department: Key Technologies
Division: Artificial Intelligence
Contact: bewerbung (at) cyberagentur.de

Artificial intelligence processes, especially deep learning, are constantly evolving and leading to increasingly complex architectures and learning processes. In safety-critical areas, however, the focus is increasingly on the traceability of AI-based procedures, which can have a significant impact on downstream decision-making processes. In order to make the outputs of complex deep learning processes more transparent and comprehensible, explainability and interpretability methods exist that examine the underlying decision patterns of an algorithm and make relevant classification features visible.

As part of a master’s thesis, various methods of Explainable AI (XAI) are to be identified, examined and compared with each other in terms of their relevance in the security sector. The resulting findings will then be transferred into a functional prototype that can be integrated into a classification process as an explainability component.

The basis of the applied part of the master’s thesis is a data set that includes image data of different classes of military vehicles and can be used for an exemplary implementation of the methods developed.

Department: Key Technologies
Division: Artificial Intelligence
Contact: bewerbung (at) cyberagentur.de

High data availability plays a crucial role in the development of machine learning models. The better this data is described and prepared, the more likely it is to obtain efficient and accurate models. In particular, data labeling, i.e. the assignment of data to a class, is a time-consuming and resource-intensive process that is used especially in supervised learning environments. In order to optimize such learning processes, label-efficient methods exist that minimize the labeling process and still enable the development of machine learning models with high accuracy. Examples of these label-efficient methods are semi-supervised learning (e.g. pseudo-labeling) or self-supervised learning (e.g. SimCLR).

The aim of a master’s thesis is to investigate the extent to which such label-efficient learning methods are suitable for application to a specific image classification problem. To this end, SOTA technologies are to be identified, compared and transferred into a functional prototype for image classification.

The basis of the applied part of the master’s thesis is a data set that includes image data of different classes of military vehicles and can be used for an exemplary implementation of the methods developed.

Department: Key Technologies
Division: Artificial Intelligence
Contact: bewerbung (at) cyberagentur.de

Generative Artificial Intelligence (GenAI) encompasses a range of generative techniques that can be used to generate data (e.g. text, images, video). Prominent representatives of GenAI are ChatGPT (Large Language Models) and Stable Diffusion (image generation).

In the context of computer vision tasks (e.g. classification, object recognition), the focus is often on the availability of extensive image material. However, there are scenarios in which this material is not available to the necessary extent (e.g. due to terrain that is difficult to access). Using generative technologies, it is possible to extend real data with synthetically generated data that is intended to increase the scope and variability of a data set in order to train classification and detection algorithms.

Generative approaches to image generation are to be investigated as part of a master’s thesis. The aim is to use a practical application scenario to identify and apply GenAI techniques that can be used to optimize computer vision tasks, especially in environments with low data availability.

The basis of the applied part of the master’s thesis is a data set that includes image data of different classes of military vehicles and can be used for an exemplary implementation of the methods developed.

Department: Key Technologies
Division: Artificial Intelligence
Contact: bewerbung (at) cyberagentur.de

In the context of machine learning, the free availability of data sets and models plays a decisive role in the commissioning and dissemination of AI-based functionalities. Against the background of this availability problem, new open source initiatives are constantly emerging in the field of artificial intelligence, in the sense of networks or platforms (e.g. Hugging Face or LAION), which set themselves the task of making data sets, source code and parameters available for use and modification.

The aim of the master’s thesis is to identify existing open source initiatives in the field of artificial intelligence and to classify them in a specially developed “Open Source AI Landscape”. Such a landscape should serve to classify and characterize the content of freely available AI modules. In addition, a procedure is to be developed with the help of which open source data as well as open source models and associated frameworks (broken down by modalities, license conditions, etc.) can be brought together and assigned to possible use cases from the cyber security context. The aim is to provide a guideline that simplifies the identification and combination of open source modules for AI-based developments and shortens the data curation and model development times associated with the development of ML solutions.

One result of the work could be the provision of a digital “Open Source AI Landscape”, which uses interactive modules (e.g. keyword search, filters, etc.) to identify freely available data sets and AI modules for a defined use case or specified framework conditions.

Department: Key Technologies
Division: Artificial Intelligence
Contact: bewerbung (at) cyberagentur.de

Embodied AI refers to a form of physically anchored artificial intelligence that learns and acts through sensors, actuators and interactions with the physical environment. While great progress has been made in traditional AI through data-driven models such as Transformers, Embodied AI faces the challenge of combining cognitive abilities with physical perception and planning actions. This opens up new possibilities for robotics, autonomous systems and human-centered applications.

The aim of this master’s thesis is to comprehensively analyze the current state of research in the field of embodied AI and to identify existing approaches, technologies and their applications. One focus will be on the integration of generative methods (e.g. large language models, large action models) and physically-informed models, which can potentially increase the self-perception and spatial perception of embodied systems. Through the use of embodied AI in versatile environments, approaches to the adaptability and learning ability of the systems in unknown environments will also be considered, as well as methods of near-real-time integration of multi-modal content (e.g. video sequences, audio, text) into the learning and execution processes. In addition, development potential and future research directions are to be derived specifically with regard to the security sector.

Department: Key Technologies
Division: Artificial Intelligence
Contact: bewerbung (at) cyberagentur.de

AI-native computing refers to a paradigm that relies on the deep integration of AI methods into all layers of IT architecture, opening up new possibilities for adaptive, autonomous and self-optimizing computing. This can create new standards in areas such as edge computing, autonomous systems and personalized digital environments. These technologies also include AI operating systems, in which artificial intelligence acts as a central control system for software, hardware and applications and provides users with applications and interfaces tailored to their needs. The first well-known representatives of such technologies are, for example, the “Computer Use” tool developed by the company Anthropic, which provides an AI model to interact with a desktop environment in a human-like manner, or AI agents for software development such as Copilot, which are intended to lay the foundation for native AI functionalities in computer architectures.

The aim of this master’s thesis is to analyze the current state of research in the field of AI-native computing and to identify development potential and future fields of application. Different topics can be focused on, such as AI-controlled operating system architectures, agentic AI systems or autonomous software ecosystems. The results of the scientific work can be overviews and evaluations of existing systems, the derivation of recommendations for action for future AI research or the production of small prototypes for demonstration purposes.

Department: Key Technologies
Division: Cryptology
Contact: bewerbung (at) cyberagentur.de

Description: Compared to traditional cryptographic methods (e.g., RSA or Diffie-Hellman), post-quantum cryptography (PQC) is a relatively new field of research. However, since it is widely recognized that the transition to PQC schemes represents a significant step forward, there is currently a great deal of research activity in this field. The aim of this project is to achieve very concrete results regarding the implementation of selected PQC schemes on a Field-Programmable Gate Array (FPGA). Of particular interest are various performance metrics (speed, power consumption, and memory requirements) as well as metrics that characterize vulnerabilities in side-channel analysis. Which of these metrics will be the focus depends on the applicant’s background and interests. The Cybera Agency will provide a suitable FPGA as well as the necessary tools for conducting (selected) side-channel analyses.

Responsibilities: The student will be responsible for the following tasks:

Initialization: Configuring an FPGA to execute selected PQC methods, either using existing software libraries or
by implementing the functionality ourselves

Track A: Conducting performance benchmarks
Track A: Optimizing the implementation
Track B: Conducting statistical analyses to identify potential
vulnerabilities (e.g., using TVLA)
Track B: Conducting side-channel attacks
Track B: Implementing techniques to defend against identified side-channel attacks

Report: Writing a report to discuss the results

Student Profile: Required: Programming experience
Required: A solid understanding of linear algebra
Required: Basic knowledge of public-key cryptography
Preferred: Some experience with a hardware description language (e.g.,
, VHDL, or Verilog)
desirable. Initial experience with PQC methods (e.g., lattice-based or
code-based)
Desirable: Basic knowledge of signal processing and/or statistical analysis

Department: Cybersecurity of Complex Systems
Division: Emergent Digital Systems for Government and the Armed Forces
Contact: bewerbung (at) cyberagentur.de

What infrastructures and architectures represent further developments in cloud technologies? — On the path to the post-cloud era.

In today’s world, cloud infrastructures have become indispensable. In both professional and personal contexts, many services are only available and usable when connected to the cloud. Dependencies on cloud providers often conflict with the concept of sovereignty. This thesis aims to explore the concept of the cloud further. What might come next? Do cloud architectures still make sense for various future scenarios, or are there better alternatives, such as decentralized structures? What are the respective advantages and disadvantages of each, and what research questions need to be answered to make this a reality?

The aim of this thesis is to identify and analyze (Bachelor’s) and to research and develop (Master’s) new, revolutionary post-cloud approaches. Various aspects are conceivable, such as network topologies, collaboration approaches, backup strategies, or other topics relevant to a post-cloud architecture.

Department: Cybersecurity of Complex Systems
Division: Emergent Digital Systems for Government and the Armed Forces
Contact: bewerbung (at) cyberagentur.de

How can we prevent malicious solutions from making it into the solution space?

In population-based optimization methods, different solutions often compete against one another to find an optimum during the course of the process. Applications can be found in the fields of routing, structural optimization, medicine, finance, and energy supply. One possible attack vector is the injection of malicious solutions that could steer an algorithm toward a false optimum, which may benefit the attacker. The solution to a problem could then contain vulnerabilities that could be exploited during implementation or keep the system’s performance at a low level.

The aim of this thesis is to explore new methods and approaches that make optimization algorithms of this kind more secure and robust against injected malicious solutions. To this end, specific attack scenarios will be developed, attack vectors defined, and solution strategies evaluated.

Department: Cybersecurity of Complex Systems
Division: Predictive and Cognitive Systems
Contact: bewerbung (at) cyberagentur.de

A key aspect of future predictive and cognitive systems is selecting the data relevant to the system’s users, making it available, and presenting it in a clear and user-friendly format. Augmented Reality (AR) represents a promising approach for implementing these requirements in practice. The subject of an internship or thesis should therefore be to examine the current state of the art in research and practice (e.g., including video games) regarding the integration of AR into complex systems, to provide a comprehensive overview, and to identify potential future developments. Of particular interest here is the user’s perspective—that is, examining the conditions under which the use of AR is helpful rather than confusing, and can thus provide genuine added value for the user.

Department: Cybersecurity of Complex Systems
Division: Predictive and Cognitive Systems
Contact: bewerbung (at) cyberagentur.de

How significant are the effects of an extreme space weather event (primarily from the Sun, e.g., a strong CME or solar storm) on global and space-based infrastructure (power grids, radio communications, the Internet, satellite-based navigation, or time synchronization) and on everyday life on Earth?

The interaction between space weather and our modern world has been known for a long time. For example, a geomagnetic storm in March 1989 caused a nine-hour power outage in Canada. Are measures already being taken in the energy sector (substations, transformers), water supply, transportation, or communications to counteract the effects of increased solar activity?

This study will place a special focus on expert assessments. The first step would be to identify these experts, such as operators of critical infrastructure or researchers. Which sectors are particularly affected? How is increased solar activity handled in practice? Is there a real threat of a global blackout in the event of unforeseen space weather events? Is there a gap between the predicted worst-case scenarios and reality? Would the damage be repairable? What might appropriate protective measures look like? Could the voltage induced by geomagnetic storms be used to generate energy?

Department: Cybersecurity of Complex Systems
Division: Critical Infrastructure Protection
Contact: bewerbung (at) cyberagentur.de

Whether in our personal lives or at work while working from home, end-to-end encrypted communication is ubiquitous. Modern communication platforms have long since moved beyond one-on-one chats to include group conversations and video conferences with multiple participants at the same time.

At the protocol level, several approaches have become established in recent years, some of which were developed specifically for secure group communication. Of particular relevance here are Messaging Layer Security (MLS), Signal’s Double Ratchet protocol, and MegOLM, which is used in the Matrix ecosystem. Currently, the Double-Ratchet protocol is considered the most widely used, but MLS has since become an IETF standard, which is why its implementation and adoption are being actively promoted. MegOLM builds on this protocol but is rarely used outside the Matrix ecosystem. One platform that relies on MegOLM in addition to Matrix is gather.town.

The aim of this thesis is to systematically compare these three protocols, identify their security guarantees, threat models, and design decisions, and evaluate their suitability for future requirements. Particular emphasis is placed on the question of how robust the respective protocols are against future developments, such as the possible emergence of quantum computers.

Department: Cybersecurity of Complex Systems
Division: Critical InfrastructureProtection
Contact: bewerbung (at) cyberagentur.de

A crisis response! We must act quickly and efficiently, but communications have broken down. Each response team can now only communicate with its immediate neighbors. Nevertheless, important information must reach its intended recipients, supplies must get to where they’re needed, and we must be able to respond to unforeseen changes. Impossible? No! If each group knows exactly how to pass on specific information, the entire operation can function like a computer that efficiently transmits messages and optimally distributes supplies without a central control center. This is based on a computational concept from theoretical computer science known as cellular automata: Several elements—the cells—are in contact with their immediate neighbors. At regular intervals, each cell evaluates the information it receives from its neighbors and then sends an appropriate response. The model is adaptable in many ways, such as the possible arrangements of cells, the degree of diversity among the cells, or whether a cell can send different signals to its neighbors in each instance. This versatility has been shown to allow for the emergent description of nearly all capabilities.

The goal now is to investigate the extent to which this computational concept is suitable for coordinating complex operations in a decentralized manner. To clarify this, both theoretical and practical questions must be answered: Can a cellular automaton be designed to distribute goods efficiently? Is this also possible if the network changes during the operation? Can an algorithm be designed to be fault-tolerant? How can such an algorithm be protected against targeted disruption? How well—or up to what level of complexity—can humans follow these rules?

Given this focus, this research question lends itself to theoretical and mathematical analysis, as well as to testing the psychological feasibility of potential approaches in studies and workshops.

Department: Cybersecurity of Complex Systems
Research Area: Emerging Disruptive Technologies
Contact: bewerbung (at) cyberagentur.de

Critical raw materials—particularly rare earth elements and strategic minerals—form the basis of numerous modern electronics, communications, and defense systems. At the same time, the European Union is highly dependent on imports from geopolitically sensitive regions, which creates vulnerabilities along the supply chains. Against the backdrop of increasing geopolitical tensions, the issues of self-sufficiency, resilience, and securing strategic raw materials are gaining in importance.

The aim of this master’s thesis is to systematically map the use of critical minerals in selected security-related electronics and communications products and to estimate future demand. To this end, existing studies, regulatory frameworks (e.g., the EU Critical Raw Materials Act), and technical specifications will be analyzed. In addition, dependencies, bottlenecks, and potential substitution options will be evaluated. The thesis may result in a prototype mapping tool that transparently illustrates the raw material dependencies of key European technologies and derives recommendations for a more resilient raw materials strategy.

Department: Cybersecurity of Complex Systems
Research Area: Emerging Disruptive Technologies
Contact: bewerbung (at) cyberagentur.de

Modern societies are increasingly dependent on complex, tightly interconnected technical systems. The failure of individual components—particularly in the area of space-based infrastructure such as navigation, communication, or Earth observation—can have far-reaching consequences that go beyond the immediate damage. However, such second- and third-round effects are often difficult to predict and have so far been inadequately accounted for in risk analyses.

This master’s thesis will examine how the failure of critical systems spreads systemically and which mechanisms lead to cascade effects. A focus may be placed on space-based services, the failure of which affects numerous civilian and security-related sectors. To this end, existing models of system interdependence will be analyzed, typical failure chains identified, and illustrative scenarios developed. The aim is to develop a methodological framework that improves the assessment of such knock-on effects and provides decision-makers with a sound basis for resilience and redundancy strategies.

Department: Cybersecurity of Complex Systems
Research Area: Emerging Disruptive Technologies
Contact: bewerbung (at) cyberagentur.de

Space software forms the backbone of modern satellite, communications, and navigation systems. At the same time, it is subject to specific requirements regarding reliability, radiation resistance, and security architecture. Despite these high standards, there are numerous known and suspected vulnerabilities, which are often documented in a fragmented manner or are difficult to access. Systematic identification and assessment of such vulnerabilities is therefore essential for identifying risks early on and developing appropriate protective measures.

The aim of this master’s thesis is to develop a structured method for collecting, categorizing, and evaluating vulnerabilities in space-related software. To this end, existing vulnerability databases, scientific publications, and security-related standards will be analyzed. Based on this analysis, an evaluation model will be developed that takes into account both technical criticality and potential impacts on mission objectives. A prototypical repository or classification scheme can serve as the result and contribute to the systematic security analysis of space-based systems.