MASTER THESES
We constantly offer M.Sc thesis projects in the areas of decentralized systems security, ML security, and platform security. If you are interested in doing your M.Sc. thesis in these areas, please contact us at inf-infsec+applications@rub.de and provide the following information:
- A short (i.e., 1-2 sentences) motivation explaining your preferred topic (e.g., decentralized systems security, ML security, TEE security). Please see below more details about our current topics. If you are unsure about your topic, just mention this in your application and we can try to find a topic that matches your background/profile.
- A concise (i.e., 1-2 sentences) description of your programming skills
- Optionally, your transcript of records (i.e., your B.Sc. / M.Sc. grades) might help us get a better idea about your profile to assign you a suitable thesis topic.
General Theses Topics for the following research areas:
Decentralized systems security is receiving considerable attention nowadays from the community as it finds direct applicability in the financial sector, in the IoT sector, among many others. The thesis addresses the problem of analyzing, implementing and evaluating a number of security primitives for existing decentralized technologies.
During the course of this thesis, the student is expected to complete the following milestones:
- Perform a survey of existing decentralized security technologies.
- Analyze and implement a selected number of technologies.
- Evaluate the performance of the devised technologies.
Advances in machine learning have enabled the widespread adoption of AI technologies, making them increasingly more present in critical decision-making processes. The massive use of ML however opens the door to unprecedented risks and challenges. The main goal of this thesis project will be to explore and address vulnerabilities of ML in adversarial settings, with focus on security and privacy. The project tasks may involve analyzing, implementing, and evaluating defensive techniques to enhance the security and privacy of ML approaches.
During the course of this thesis, the student is expected to complete the following milestones:
- Survey existing vulnerabilities of ML algorithms from a security/privacy viewpoint.
- Analyze and implement selected technologies to mitigate vulnerabilities of ML.
- Evaluate the performance of the devised technologies.
Requirements:
- Intermediate familiarity with (adversarial) machine learning
- Intermediate programming experience in Python
- Optional: Experience with Python ML framework PyTorch / TensorFlow
Digital platforms bring innovation opportunities by connecting physical and digital objects with the cloud. This also opens up new adversarial opportunities. Trusted Execution Environments (TEEs) may mitigate such threats, but they require careful design to meet performance, flexibility and security required by online applications. This thesis focuses on implementing and evaluating TEE-based solutions for platforms.
During the course of this thesis, the student is expected to complete the following milestones:
- Perform a survey of existing Trusted Execution Environments.
- Analyze and implement a selected number of technologies.
- Evaluate the performance of the devised technologies.
- Ability to program in C/C++/Rust
- Familiarity with Python
- Knowledge of software security
- Optional side-channel attack knowledge