Wi-Fi Traffic Classification Using Time Series Data [Duration 1-2 terms]

Suggested Team Size: 2-3 students 

Duration: 1-2 semesters

Project Description:

In this traffic classification project on WiFi networks, the goal is to accurately identify and categorize traffic into specific classes such as gaming, live streaming, and web browsing. The data collection process involves capturing time series data from WiFi access points, including packet loss, PHY rates, channel utilization, and signal strength metrics. This raw data is then preprocessed and labeled according to the type of traffic being observed. Once the dataset is prepared, machine learning models are deployed to classify the traffic types. Popular models like Random Forest, Boosting algorithms, and Neural Networks are trained to recognize patterns in the time series data, enabling traffic classification and optimization of network resources based on traffic type. 

- Suggested Reading List: https://www.researchrabbitapp.com/collection/public/JLNNRD9PLN

- https://patents.google.com/patent/US20240028670A1/en