Research
My research is primarily focused on decentralized distributed systems and machine learning, specifically the field of decentralized federated learning. The combination of these topics poses a particularly interesting possibility for the development of robust and privacy-preserving machine learning.
Thesis Topics
Topics for supervision of theses are primarily in my fields of research:
decentralized distributed systems, online privacy, decentralized machine learning.
In these areas we occasionally find interesting questions
that are suitable to elaborate in a thesis.
If you have an idea for a problem in these fields
that might be interesting to work on,
we can work out the specifics of the thesis together.
Aside from that,
I have also supervised theses on security questions on occasion.
However, since my primary research interests lie in other areas,
I cannot oversee topics that require intimate knowledge of specific methods.
You can use the list of previously supervised topics below
as insipration to find a problem to work on.
Supervised Theses
- Applying security principles in NFC based employee hour tracking systems (Bachelor, September 2020)
- Evaluation of Bootstrapping Mechanisms for Persistent Streams (Bachelor, January 2021)
- Building a Stateless Privacy-Focused Microblogging Service on a Decentralized Architecture (Master, July 2021)
- Random Numbers in Cryptography - Analysis on the Prevalence of Unsafe (Pseudo) Random Numbers in Open Source Software (Bachelor, July 2021)
- Efficient Transparent Proxies for QUIC Frames in Persistent Streams (Bachelor, September 2021)
- Development of an Azure IoT Edge Module for Physik Instrumente Devices (Bachelor, June 2022)
- An Analysis of Known Security Flaws in Building Automation Protocols (Bachelor, July 2022)
- Reducing Server Load by Increasing the Efficiency of Push Notifications (Bachelor, August 2022)
- Discovery and Distribution Mechanisms for Longevity Digital Twins (Master, July 2023)
- Integrating Graphical Output Capabilities into the MIC-1 (Bachelor, September 2023)
- Design of a Customizable Lossless Data Compression Algorithm (Master, January 2024)
- Planning, execution, and evaluation of load tests of a distributed backend system for the provision of telephony functions (Master, October 2024)
- Load Balancing for Dynamic Consistency Guarantees in Web Requests (Bachelor, January 2025)
- Free-Riders in Decentralized Federated Learning: Evaluating Solutions from Peer-to-Peer Systems (Master, January 2025)
- Comparing Performance Impacts of DNS Security Mechanisms (Bachelor, February 2025)
Publications (6)
2024 (1)
-
BitTorrent-based Gossip LearningIn: 14th International Conference on the Internet of Things (IoT 2024), November 19--22, 2024, Oulu, Finland
2023 (1)
-
A Privacy-Protecting Step-Level Walking Direction Detection Algorithm Based on Floor VibrationIn: IEEE Sensors Journal
2021 (2)
-
Persistent Streams: The Internet With Ephemeral StorageIn: Tagungsband des FG-BS Herbsttreffens 2021
-
An Algorithm for Dependency-Preserving Smart Home UpdatesIn: PerIoT 2021: 5th International Workshop on Mobile and Pervasive Internet of Things (PerIoT 2021)
2020 (2)
-
Requirements and Mechanisms for Smart Home UpdatesIn: SENSYBLE 2020, 2nd workshop on Smart Systems for Better Living EnvironmentsDOI: 10.18420/inf2020_91
-
Group-based Memory Management in FyrIn: Tagungsband des FG-BS Herbsttreffens 2020
Teaching
WiSe 24/25
SoSe 2024
WiSe 23/24
SoSe 2023
WiSe 22/23
SoSe 2022
WiSe 21/22
SoSe 2021
WiSe 20/21
SoSe 2020
Oskar Carl
Research Associate
Contact
- Phone: TBD
- E-Mail: oskar.carl[@uni-due.de]
- ORCID iD: https://orcid.org/0000-0002-6042-6009
Postal address
- Building BC, Room 312
- Bismarckstr. 90 47057 Duisburg, Germany
Consultation hour
By appointment