Crowd management is of extreme importance for one-time-events (e.g., FC Bayern match) as well recurrent gatherings (e.g., U-Bahn stations in the early morning). More importantly, the recent pandemic outbreak made contact tracing even more important to mitigate its effect on our society. Knowing their flocking patterns allows for optimized resource allocation in real-time. Building an efficient solution to this problem is the main goal of this thesis with the concomitant (and inevitable) privacy concerns.
Building on existing work, this thesis will use 802.11 management frames to passively identify nearby mobile devices. Your tasks will be to build, test and evaluate:
- Time and frequency domain analysis of management frames
- Energy consumption when sampling these frames, possibly optimizing for longer battery life
- Perform a large scale studies in order to evaluate the data being collected
The main challenges of this thesis are:
- Build robust models to deal with the data used while preserving the privacy of its subjects
- Build a testbed which will be used to systematically collect large amounts of data, which would be central to the evaluation of this thesis
- (Good!) Experience with python, R or Julia
- Familiarity with C/C++
- Experience with Linux, microcontrollers and single board computers
Good to have
- (Good!) Background in Machine Learning
- Familiarity with Pub/Sub systems, noSQL databases, data visualization libraries/frameworks
If you are interested, please send me an email with your CV and transcripts (and a short introduction).
Leonardo Tonetto <tonetto at in dot tum dot de>