Looking into WiFi Channel State Information (CSI): From Signal Processing Foundations to Practical Applications
Dr Ruiqi Kong, The Chinese University of Hong Kong

Abstract: Channel State Information (CSI) offers a fine-grained view of the wireless medium that extends far beyond its conventional role in communication. It inherently contains a mixture of multipath propagation, hardware distortions, and motion information, presenting significant opportunities for cross-domain applications. In this talk, I will present my research on harnessing this potential through a fusion of signal processing and deep learning. The focus will be a deep dive into two projects on physical-layer fingerprinting, where I demonstrate how to extract device-specific hardware imperfections from CSI to generate robust RF fingerprints for accurate device authentication and identification. Furthermore, I will briefly introduce our work in WiFi sensing, which exploits delay-domain signal processing to recover motion information. Collectively, these efforts establish a comprehensive model-guided framework for translating raw CSI data into practical solutions for wireless security and sensing.

Speaker's Bio: Ruiqi Kong is an incoming Postdoctoral Fellow in Information Engineering at The Chinese University of Hong Kong (CUHK). She recently completed her Ph.D. at CUHK under the supervision of Professor He (Henry) Chen. Her research focuses on the advanced analysis of wireless signals to tackle challenges in wireless security and sensing.