Project / 2026-09-01
AI-Enabled WiFi Sensing & RF Classification
Ingesting raw RF time-series IQ data from a USRP B210 to detect motion and classify targets.
End-to-end signal processing and deep learning pipeline designed to classify human presence and motion through RF perturbations. Ingests raw IQ time-series data captured via a USRP B210 software-defined radio, transforms channel state information into spectro-temporal feature maps, and trains neural models for real-time target inference.