CSI-based Wi-Fi sensing is a passive sensing paradigm that leverages Channel State Information (CSI) — fine-grained physical layer measurements of how a Wi-Fi signal propagates through an environment — to infer properties of that environment or the people within it, such as human activity, presence, or movement, without requiring dedicated sensors or wearable devices. It matters because it enables ubiquitous, privacy-preserving sensing using existing Wi-Fi infrastructure, making it a compelling approach for applications in healthcare, smart homes, and security. Key variants are defined by the sensing task (e.g., human activity recognition, gesture detection, localization), the underlying Wi-Fi standard employed (e.g., IEEE 802.11n, 802.11ac, or the more recent 802.11ax/Wi-Fi 6 with its expanded PHY features), and the reliance on static versus dynamic CSI components, with the latter distinction having significant implications for temporal generalization and real-world deployment robustness.
Source Papers
- A survey on CSI-based Wi-Fi sensing datasets and models with a focus on reproducibility ↗ — A survey on CSI-based Wi-Fi sensing datasets and models with
- Exposing the CSI: A Systematic Investigation of CSI-based Wi-Fi Sensing Capabilities and Limitations ↗ — Exposing the CSI: A Systematic Investigation of CSI-based Wi
- Time matters: Empirical insights into the limits and challenges of temporal generalization in CSI-based Wi-Fi sensing ↗ — Time matters: Empirical insights into the limits and challen