In this project, we aim to design a program that collects CSI from a Wi-Fi router, and recognize seven actions (Circle, Clap, Slide, Kick, Push, Stand up, and Sit) by training a CNN model based on the collected data.
Instead of transmitting special packets continually, we use the Wi-Fi beacon as our Tx signal.
This project is the term project for Wireless Networking - Fundamentals and Applications class in Fall 2021. For more details, please refer to the project report.
- Tx: Asus RT-N66U
- Rx: Raspberry Pi 4B
- Server: Laptop (Macbook pro)
- Tx broadcasts the Wi-Fi beacons to the Rx.
- Rx only captures the Wi-Fi beacon packets.
- When capturing some amounts of the packets, save these packets to a file.
- Transmit this file to the edge server and get the gesture recognition result.
- nexmon-csi: Channel State Information Extraction on Various Broadcom Wi-Fi Chips
- CSIKit: CSI processing tools