WebRTC Data Collection and Analysis on NetUnicorn
Published:
The WebRTC data collection and analysis project at UCSB SNL led by Prof. Arpit Gupta
- Developed a Python and Selenium-based automated Google login system to bypass bot detection and enable automatic participation in online conferences
- Containerized data collection pipeline with Docker into NetUnicorn, increasing deployment speed by 60%
- Analyzed 10+ key indicators by reviewing 1,000+ pages of WebRTC documentation and aligning with RFC standards using NumPy, pandas, SQLite, and Matplotlib
- Automated headless ARM64 Raspberry Pi systems for data collection, efficiently aggregating 7 TB across 30 nodes concurrently