https://github.com/giovannibaccichet/nmda-derandomization
Wi-Fi Probe Request MAC Address De-Randomization Lecture ー Network Measurement and Data Analysis
https://github.com/giovannibaccichet/nmda-derandomization
Last synced: about 2 months ago
JSON representation
Wi-Fi Probe Request MAC Address De-Randomization Lecture ー Network Measurement and Data Analysis
- Host: GitHub
- URL: https://github.com/giovannibaccichet/nmda-derandomization
- Owner: GiovanniBaccichet
- Created: 2024-03-11T13:57:40.000Z (about 1 year ago)
- Default Branch: main
- Last Pushed: 2024-12-03T15:55:35.000Z (6 months ago)
- Last Synced: 2025-01-30T07:12:55.655Z (4 months ago)
- Language: Jupyter Notebook
- Homepage:
- Size: 2.64 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
Wi-Fi Probe Request De-Randomization
Network Measurement and Data Analysis Lab
Prof. Alessandro Redondi
# Lecture
This lecture delves into MAC Address de-randomization within the framework of Wi-Fi Probe Request frames transmitted during the network discovery phase of Wi-Fi devices. The accompanying repository houses notebooks and scripts tailored for the following tasks:
1. **Extraction** of data from `.pcap` format to `.csv`.
2. **Examination** of frame contents to identify pertinent patterns.
3. Implementation of **clustering**, utilizing a toy example to illustrate the significance of discerning crucial features in distinguishing between devices.
4. **Enumeration** of the total devices present in the dataset.
5. **Evaluation** of performance through clustering metrics applied to labeled data.
# Challenges
Implement a better clustering strategy, based on a threshold of equal features between Probe Request Bursts (i.e., sharing $N$% of features). Do **not** use ML models, such as K-Means, DBSCAN, etc.
Then perform MAC Address de-randomization for the devices in `data/challenge-dataset/` at various dataset sizes and for different thresholds of feature equality (e.g., $10, 20, 30, ..., 100$). Analyze the results and observe any patterns or trends in clustering metrics. What can you notice?
Finally, try to perform de-randomization on the unlabelled dataset in `data/unlabelled-challenge.csv`. How many devices are there?
> *Hint:* Observe the clustering metrics for different thresholds and dataset sizes from the first half of the challenge. Apply the insights gained to successfully solve the second half of the challenge.