Dataset Generation for ML Personal Tracker Detection
MA
State: completed by Ramize Abdili, Stefan Richard SaxerPublished: 2023-06-04
This work focuses on the collection of BLE packets through passive sniffing to generate a large dataset which can be analyzed through ML approaches. The aim of this work is to collect a large enough dataset through sniffing to see if there are any patterns in the data that could identify specific BLE devices, such as AirTags.
This work includes:
- Capturing of BLE packets over extended period
- Preprocessing and creation of the dataset
- Evaluation and consideration of appropriate ML approach
Helpful Sources:
- https://petsymposium.org/2020/files/papers/issue1/popets-2020-0003.pdf
- https://arxiv.org/pdf/1904.10600.pdf
- https://ieeexplore.ieee.org/abstract/document/8638232
- https://ieeexplore.ieee.org/document/8795319
- https://arxiv.org/pdf/2211.01963.pdf
- https://scholar.afit.edu/cgi/viewcontent.cgi?article=6008&context=etd
40% Design, 40% Implementation, 20% Documentation
Supervisors: Katharina O. E. Müller
back to the main page