Bachelor | Master | PhD – Thesis Supervision

RESEARCH AREAS AND TOPICS

STUDENTS INTERESTED to do an INTERNSHIP or pursue a THESIS (B.Sc., DI, M.Sc., or Ph.D.) are advised to directly CONTACT OUR RESEARCHERS as listed below.

Security and Privacy in Machine Learning

Below is a list of topics related to security and privacy aspects of Machine Learning / Data Mining / Artificial Intelligence. If you are interested, please contact Rudolf Mayer

Adversarial Machine Learning:

A good overview talk (in German) into Adversarial Machine Learning is given by Konrad Rieck: “Sicherheitslücken in der künstlichen Intelligenz”: https://www.youtube.com/watch?v=j6nRPg2Y62w

Adversarial Inputs (resp. robustness against adversarial inputs)
– Intro papers: “Explaining and Harnessing Adversarial Examples” (https://arxiv.org/abs/1412.6572), S&P 2017 paper on “Towards Evaluating the Robustness of Neural Networks” (https://arxiv.org/abs/1608.04644, “Making machine learning robust against adversarial inputs” (https://dl.acm.org/citation.cfm?id=3134599)
– Video: https://www.youtube.com/watch?v=yIXNL88JBWQ)
– Goal: systematically analyse existing evasion attacks and defenses and develop new attacks/defenses in specific application domains (such as industrial productions systems (https://www.sqi.at)).
Backdoor (data poisoning) attacks & defenses
– Intro papers, e.g. “Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning” (https://arxiv.org/abs/1712.05526) or “BadNets: Identifying Vulnerabilities in the Machine Learning Model Supply Chain” (https://arxiv.org/abs/1708.06733)
– Video: a talk about one defense mechanism: https://www.youtube.com/watch?v=krVLXbGdlEg
– Goal: analyse and evaluate attack vectors for poisoning attacks, evaluate their effectiveness and side-effects, as well as existing defenses, and develop new detection/defense mechanisms
Membership inference attack
– Intro papers: “Membership Inference Attacks against Machine Learning Models” (https://arxiv.org/abs/1610.05820)
– Video: Reza Shokri, “Membership Inference Attacks against Machine Learning Models”: https://www.youtube.com/watch?v=rDm1n2gceJY
– Goal: analyse and evaluate attack scenarios for membership inference, analyse existing attack and defense patterns, and develop new mechanisms
Other attacks e.g.
– Model stealing (Paper “Stealing Machine Learning Models via Prediction APIs” (https://www.usenix.org/conference/usenixsecurity16/technical-sessions/presentation/tramer,
– Video: http://0b4af6cdc2f0c5998459-c0245c5c937c5dedcca3f1764ecc9b2f.r43.cf2.rackcdn.com/sec16/tramer.mp4)
– Model inversion: Paper “Model Inversion Attacks that Exploit Confidence Information and Basic Countermeasures” (https://www.semanticscholar.org/paper/Model-Inversion-Attacks-that-Exploit-Confidence-and-Fredrikson-Jha/721c4169f2f81d2f0b475381a5af46589d36ef92)

Privacy-preserving Machine Learning/Data Mining

Privacy-preserving analysis of data is becoming more relevant with the increasing amount of personal data being gathered. Several different approaches aiming at this problem exist, e.g.:

Privacy-preserving data publishing
– k-anonymity, l-diversity, etc.
– Differential privacy, including local differential privacy
– Synthetic data generation
– Goal: evaluation of privacy protection, utility of the published data, novel attack mechanisms, application of differential privacy to machine learning models, …
Privacy-preserving computation
– Secure Multi-party computation (SMPC / MPC). Teaser video: https://www.youtube.com/watch?v=l25jcolQW6Q, more detailed explanation: https://www.youtube.com/watch?v=NOtsxHoIcWQ
– Homomorphic encryption. Intro Video: https://www.youtube.com/watch?v=f-85Tmr7ZWY, https://www.youtube.com/watch?v=xlcb_G1_rzk
– Federated learning (e.g. https://federated.withgoogle.com/)
– Goal: evaluation of effectiveness (e.g. accuracy) and efficiency of privacy-preserving approaches, compared to a base line of centralised learning. Application of approaches to new algorithms, data types, etc.

Watermarking / fingerprinting of datasets

Goal: evaluation of schemes for their robustness of attacks, vs. their data utility, e.g. measured by effectiveness in machine learning tasks

Digital Sustainability and Resilience

Demonstrate a complex yet resilient component based infrastructure by implementing a (Netflix) Chaos Monkey demonstration setup with Docker/Kubernetes components. Edgar Weippl
Runtime analysis of selected open source system(s): use debug interface of (Java) virtual machine to analyse runtime behaviour, draw a “code-landscape” and identify processes. Edgar Weippl
Analyse the complexity of selected Open Source projects following established complexity- and entropy-indicators. This might include static code analysis and analysis of code changes (source code repository). Edgar Weippl
Implement a dynamic simulation of an organisation and deduce efficiency “bottle-necks” using queuing theory. Edgar Weippl

Security for Cyber Physical Production Systems

Network Security in CPPS. Which tools (like metasploit) are readily available for pentesting CPPS (instead of business systems)? Edgar Weippl
Forensics Analysis in CPPS. Which tools exist that can be used to analyse CPPS? Are there guidelines on how to proceed? Edgar Weippl
Secure Software Engineering Processes in CPPS. Edgar Weippl

Combinatorial Testing for Information Security

Test case generation for security testing: Dimitris Simos
Combinatorial testing for hardware malware detection: Dimitris Simos
Linux kernel testing for security bugs: Dimitris Simos
Security protocol testing: Dimitris Simos
Web application security testing: Dimitris Simos
Combinatorial Testing of Tor Network: Dimitris Simos
Mathematical modelling and analysis of security vulnerabilities: Dimitris Simos
Optimizations Algorithms for Combinatorial Testing: Dimitris Simos
Discrete Structures for fault-localization in Combinatorial Testing: Dimitris Simos

Cryptocurrencies

Bitcoin scalability simulations: Aljosha Judmayer
P2P networking aspects of cryptocurrencies: Aljosha Judmayer
Alternative consensus/mining methods: Aljosha Judmayer
Alternative applications and designs of block chain based consensus Systems: Aljosha Judmayer
Bitcoin wallet Security: Aljosha Judmayer
Altcoin evaluation and categorisation: Aljosha Judmayer

Cryptography

Cryptographic engineering for security applications: Dimitris Simos
Hard problems for post-quantum cryptography: Dimitris Simos
Mathematical properties for testing implementations of crypto schemes: Dimitris Simos
Randomness extractors from discrete structures: Dimitris Simos
Design Estimators for quantum protocols: Dimitris Simos

Fileformat Security

File-format hacks (such as funky file formats): Edgar Weippl
Hiding executable code in data files (such as Word, etc.): Edgar Weippl

Internet of Things (IoT) Security and Privacy

Update mechanisms for constrained IoT systems
Secure embedded operating systems for IoT applications: Georg Merzdovnik
IoT network protocol security analysis
Privacy leaks and protections for IoT devices and protocols
Lightweight intrusion detection for IoT systems: Georg Merzdovnik
Efficient implementation of security and cryptography for IoT
Moving-target defenses in IoT environments
Resilient communication protocols for IoT

Internet of Things (IoT) – Constrained Networks

Integration of RIOT OS into SecureWSN
– TinyIPFIX for Data Gathering
– Aggregation Support
Implementation of security solutions
Optimization of system lifetime using energy harvesting mechanisms
Extention of visualization mechanisms
Integration of warning system for unnormal behavior (e.g., wrong values or to high packet loss rate)
Integration of “Trust-Check” for devices
Establishment of bidirectional communication (e.g., for updating purposes of intervals)

Topic description available under https://corinna-schmitt.de/student-theses.html

Cyber-Physical System Security

Power consumption modulation measurements (GPU, CPU, disk access, etc.): Johanna Ullrich
Attacks on cyber-physical systems: Johanna Ullrich
POSIX blackbox live forensic: Aljosha Judmayer

Mobile Security and Privacy

Dynamic Instrumentation: Georg Merzdovnik
Static code deobfuscation: Georg Merzdovnik
Android native code analysis: Georg Merzdovnik
Mobile OS Security (Firefox OS, Tizen,…): Georg Merzdovnik
On-Device application analysis: Georg Merzdovnik

Network Security

IPv4/IPv6 sibling detection: Johanna Ullrich
Long term observations of IPv6 addresses: Johanna Ullrich
Collection of public IPv6 address data sets: Johanna Ullrich
Internet cartography: Johanna Ullrich
Development of covert and side channels: Johanna Ullrich
IPv6-based network scanning and reconnaissance: Johanna Ullrich
IPv6 address formats (analysis and redevelopment): Johanna Ullrich
Perception of IPv4 vs. IPv6 addresses (user study): Johanna Ullrich
Analyzing top Internet page lists e.g., Alexa: Wilfried Mayer
Net neutrality measurements: Wilfried Mayer
RIPE Atlas data analysis: Wilfried Mayer
Transparency of Internet-based protocols: Wilfried Mayer

Privacy Enhancing Technologies

Tor network measurements: Wilfried Mayer
Tor relay operator transparency: Wilfried Mayer
Impact of Certificate Transparency on non-HTTPS-based ecosystems: Wilfried Mayer
Censorship measurement technologies: Wilfried Mayer

Cloud Security

Modelling of networking capabilities of hypervisors (Xen, KVM, etc.): Johanna Ullrich
Modelling of networking capabilities of containers: Johanna Ullrich
Development of covert and side channels: Johanna Ullrich
Network cartography in clouds: Johanna Ullrich
Network attacks in clouds: Johanna Ullrich

Research in Information Security

Comparing military research and academic research – analysis of de-classified content: Edgar Weippl
Analysis of security in standardization processes: Johanna Ullrich